ORIGINAL

Niger J Paed 2015; 42 (3):194 –198

Kayode-Adedeji BO

Owa JA

Akpede GO

Alikah SO

Evaluation of Jaundice meter in the

assessment of jaundice among

Nigerian preterm neonates

DOI:http://dx.doi.org/10.4314/njp.v42i3.4

Accepted: 24th February 2015

Abstract: Background: The ob-

jective assessment of the severity

of neonatal jaundice is Total Se-

rum Bilirubin (TSB) determina-

tion, which requires multiple

blood sampling. This has inherent

problems, including risks of anae-

mia and infection. Transcutaneous

Bilirubinometry (TcB) is a reli-

able, non-invasive alternative,

however there is paucity of data

on its performance in black pre-

term neonates.

Objectives: To evaluate the corre-

lation between transcutaneous

bilirubin (TcB) and total serum

bilirubin (TSB) among Nigerian

preterm neonates, and to deter-

mine the parameters affecting the

relationship.

sponding TcB levels were obtained

using Jaundice Meter (JM-103).

The neonates were stratified into

gestational age and birth weight

groups.

Results: A total of 189 paired TcB

and TSB levels were obtained

from 60 neonates. The

Mean (sd) TcB level of 11.4 (3.1)

mg/dl was significantly higher

than the mean TSB level of 10.2

(2.8)mg/dl (p= 0.028). The overall

correlation coefficient between

TcB and TSB was 0.98 and it was

not significantly affected by the

gestational age, birth weight and

bilirubin levels.

Conclusion: Transcutneous

bilirubin strongly correlates with

total serum bilirubin levels among

Nigerian preterm neonates, irre-

spective of gestational age, birth

weight and the degree of jaundice.

(

)

Kayode-Adedeji BO

Akpede GO, Alikah SO

Department of Paediatrics,

Irrua Specialist Teaching Hospital,

Irrua, Edo State, Nigeria.

Email: idinovsky@yahoo.com

Owa JA

Department of Paediatrics and Child

Health, Obafemi Awolowo University,

Ile-Ife, Osun State, Nigeria.

Method: Jaundiced preterm neo-

nates delivered between 28 and 36

weeks of gestation admitted

at the Irrua Specialist Teaching

Hospital (ISTH), Nigeria were

recruited . The TSB

levels were determined by spec-

trophotometry while the corre-

Keywords: hyperbiliruinaemia,

jaundice meter, Nigerian,

preterm, neonates, transcutaneous

bilirubinometry

Introduction

The Jaundice meter is portable and rechargeable. More-

over, gestational age and skin pigmentation have been

16

Neonatal Jaundice (NNJ) is a major cause of neonatal

morbidity and mortality in de_-v₃eloping

speculated as potential limitations in the use of TcB .

Previous studies among term neonates in Nigeria sug-

gested that skin pigmentation may not constitute a major

limit_,a₁₇tion to the use of transcutaneous bilirubinome-

1

countries, including Nigeria . Jaundice is more preva-

lent, more severe and associated with a higher risk of

bilirubin encephalopathy, and runs a more protracted

course in preterm than term neonates

1

ters . There is however a dearth of data among preterm

4

-7

.

neonates, especially among Blacks in which transcutane-

ous bilirubinometry may be of particular benefit. This

study set out to evaluate the correlation of the Jaundice

Meter (JM-103) with TSB in assessing the severity of

jaundice in Nigerian preterms and the factors affecting

its correlation.

Transcutaneous bilirubinometry, a non-invasive method

of assessment of severity of neonatal jaundice (NNJ) has

been extensively studied in term infants and shown to

significantly reduce the need for blood sampling an_-₁d₅ the

8

delay in intervention for NNJ in the Caucasians . In

contrast, less has been done on the reliability of transcu-

taneous bilirubinometry in preterm neonates, particu-

larly in the Blacks. Transcutaneous bilirubinometry may

be more beneficial to the Blacks in the gross resource-

limited developing countries, particularly in Africa, be-

cause of the unfavourable conditions under which NNJ

is managed. These conditions include poor access to

healthcare facilities, erratic electricity, inadequate diag-

nostic and treatment facilities.

Subjects and Method

The study was carried out at the Special Care Baby Unit

(SCBU) and Post Natal Ward (PNW) of Irrua Specialist

Teaching Hospital (ISTH), Edo State, Nigeria. Approval

1

95

was obtained from the Research and Ethics Committee

of the hospital while written informed consent was ob-

tained from the parents of the neonates involved. Sub-

jects were consecutive jaundiced preterm neonates be-

tween the gestational ages of 28 and 36 weeks. They

were recruited when the determination of serum

bilirubin levels was indicated based on clinical assess-

ment. Gestational age was assessed on admission using

the results of antenatal ultrasonography, Naegele’s rule

coefficients of sternal and forehead TcB readings with

TSB level was not statistically significant(0.98 versus

0.97, p = 0.76). The data obtained from sternal TcB

readings in relation to TSB levels are subsequently pre-

sented below.

The mean (sd) TcB level of 11.4 (3.1)mg/dl was signifi-

cantly higher than the mean TSB level of 10.2 (2.8)mg/

dl (p= 0.028).Of the 189 pairs, the TcB readings were

higher than the

13

and clinical assessment using the new Ballard Score .

Neonates who had undergone exchange blood transfu-

sion, and those whose gest₈a_,t₁i₀onal ages could not be

determined were excluded . Neonates on phototherapy

who were otherwise eligible for recruitment were in-

cluded; an opaque plaster was applied to the site used

for Tc Breadings, to prevent the bleaching effect of pho-

TSB levels in 181 (95.8%), lower in four (2.1%) and

equal in four (2.1%) cases. The difference between TcB

and TSB was less than 1 mg/dl in 134 (70.9%), 1 - 1.9

mg/dl in 40 (21.2%) and 2 - 2.8 mg/dl in 15 (7.9%) of

the paired readings.

Fig 1 shows the linear regression of TcB readings on

TSB levels. There was a strong positive

18,19

totherapy on the skin

.

relationship between TcB readings and TSB levels, with

correlation coefficient (r) of 0.98 and coefficient of de-

2

The TcB readings were carried out using the Jaundice

Meter-103 (JM-103). The JM-103 has been designed to

overcome the limitations of older versions of Jaundice

meters with respect to gestational age and skin pigmen-

termination (R ) of 96 (p < 0.01). The linear relationship

between the two is described by the equation TSB =

TcB x 0.93.

10

tation .₈T_,₁₀he readings were taken over the forehead and

sternum . Sampling for TSB was through venepunc-

ture immediately after TcB reading and analysis for TSB

was done at the Chemical Pathology Laboratory of the

hospital, using an SM23A spectrophotometer (Malloy

and Evelyn Method). Sampling was based on clinical

indication; an average of three samples was obtained

from each baby. The neonates were stratified into gesta-

tional age groups of 28-30 weeks, 31-33 weeks and 34-

Fig 1: Linear regres-

sion of TcB on TSB

for all the readings.

3

1

6weeks and into birth weight groups less than 1000gm,

000-1499gm, 1500-2499gm and2500gm and above.

The effects of total serum bilirubin levels, gestational

age (GA) and birth weight on the relationship between

TcB readings and TSB levels.

Means of TcB readings and TSB levels were compared

using paired Student’s t-test while the correlation coeffi-

cient (r) between TcB readings and TSB levels was de-

termined using linear regression analysis. The differ-

ences in means of TcB readings and TSB levels among

the various categories of preterm neonates were com-

pared. P values less than <0.05 were considered to be

statistically significant.

Eleven neonates had TSB levels ≥15 mg/dl, the correla-

tion coefficient between TcB and TSB in

this TSB range was 0.78 while the correlation coeffi-

cient was 0.98 at TSB levels ≤ 14.9mg/dl. The differ-

ence clinically significant, although it did not attain sta-

tistically significance (p = 0.09) probably due to the

small number (n=11) in that range. The mean (SD) of

the difference between TcB readings and TSB levels of

Results

1

.16 (1.02) mg/dl at TSB levels ≥15 mg/dl was signifi-

cantly higher than that of 0.76 (0.62)mg/dl at lower TSB

levels ≤ 14.9mg/dl (p = 0.047).

A total of 189 paired measurements of TcB readings and

TSB levels were obtained from 60 neonates, comprising

3

5 (58.3%) males and 25 (41.7%)females. They were all

The distribution of the means (SD) of TcB readings,

TSB levels and differences in TcB-TSB in

born to Nigerian parents. The gestational age ranged

from 28 to 36with mean (SD) of 31.5 (2.7) weeks while

the birth weight ranged between 900 and 2600 with

mean (SD) of 1624 (583) gm. Twenty-eight (46.6%) of

the neonates were 28-30 weeks, and 16(26.7%) each in

the 31-33 weeks and 34-36 weeks categories.

relation with GA are shown in Table 1. The means of

TSB readings and TcB levels increased significantly

with GA (p< 0.01). There was however no statistically

significant difference in the means of TcB readings and

TSB levels in any of the GA categories. The correlation

coefficients ‘r’ among the GA of groups 28-30 weeks,

Overall relationship between TSB and TcB levels

3

0

1-33 weeks and 34-36 weeks were (r =0.97, 0.98 and

.98, respectively). There was no statistically significant

There was no statistically significant difference between

the mean of the sternal and forehead TcB readings (11.4

versus 11.3 mg/dl;95% CI of the difference = -0.51,

difference in the correlation coefficient (r) within the

three GA groups.

0

.71, p= 0.746). The difference between the correlation

1

96

,17

and pre-

term Caucasian neonates may be due to the improved

sensitivity of the study instrument, the JM-103, which

was specifically designed to overcome the limitations of

the older versions of jaundice meters_,₁₀with respect to

1

Table 1: Table showing the comparative statistics of TSB and

TcB for different gestational age categories.

previous studies on African term neonates

Gestational

age cate-

gory

No of

readings

Mean

(SD)

TSB(mg/

dl)

Mean

(SD)

TcB(mg/

dl)

Mean

difference

(95% CI)

t-value

(p-value)

(

weeks)

8

28-30

88

8.1 (1.8)

8.5 (1.8)

(-1.4, 0.6)

-0.83 (0.41)

gestational age and skin pigmentation.

3

All babies

P values

1-33

4-36

49

52

189

11.1 (1.1)

12.8 (2.3)

10.2 (2.8)

< 0.01

11.8 (2.0)

13.4 (2.4)

11.4 (3.1)

< 0.01

(-1.9, 0.5)

(-2.3, 1.1)

(-2.3,0.2)

< 0.01

-1.23 (0.23)

-0.72 (0.48)

-1.79 (0.15)

The tendency of higher TcB readings than TSB levels in

the prese₈n_,₂t₀_,s₂t₃udy is consistent with the findings of other

workers.

This tendency for the TcB readings to be

Key: S.D = standard deviation; TSB = total serum bilirubin; TcB =

transcutaneous bilirubin; CI = confidence interval

higher than the TSB levels is not likely to result in major

errors in the management of preterm neonates with jaun-

dice. This is because at least one value of TSB is usually

obtained before instituting treatment.

Table 2 shows distributions of means (SD) values of

TcB and TSB, and the means (SD) of the differences

between TcB and TSB levels in relation to birth weight

categories. The Table shows the comparison between

the means of TcB and TSB for each of the birth weight

category. In each birth weight category, the mean of

TcB was higher than that of TSB, though the difference

was not statistically significant (p = 0.09 for < 1000gm,

p = 0.51 for 1000-1499gm, p = 0.25 for 1500-2499gm

and p = 0.42 for ≥ 2500gm. The means of both TSB and

TcB increased significantly with birth weight (p <0.01).

The correlation coefficients were r = 0.85 (p < 0.01); r =

Moreover, the practical implication is a probable slight

increase in the rate of interventions with phototherapy, if

only the TcB readings were used for the assessment of

the severity of the jaundice and in taking decision.₈_,₂₀

In agreement with the results of previous studies, the

correlation coefficient between TcB readings and TSB

levels in the present study was higher with TcB meas-

urements taken over the sternum than that taken over the

forehead, although the difference was not statistically

significant. The difference has been attributed to the

probable bleaching effect of light on the fore₂h₃ead com-

pared to the relatively less exposed sternum, although

precautions were taken to minimize this effect. The ster-

num is thus the preferred site for TcB measurements.

The present study also revealed a lower correlation coef-

ficient at₈h_,₁i₁g_,₁h₇er TSB values as had been previously

reported.

0

1

.98 (p < 0.01); r = 0.97 (p < 0.01); and r = 0.98 (p <

.01) for the weight categories of <1000gm, 1000-

499gm, 1500-2499gm and ≥2500gm respectively. The

weakest correlation coefficient between TcB and TSB

was thus in the <1000gm weight category. However

there was no statistically significant difference between

the correlation coefficients (r) between the various birth

weight categories (p = 0.08).

The recommendation for the use of TcB as a screening

device is based on the assumption that serum and tissue

bilirubin levels are in constant equilibrium. This may

not be totally true. It is possible that at higher serum

bilirubin levels, the rate of equilibration between the

serum bilirubin and tissue bilirubin is reduced and more

bilirubin is retained in the tissues, including subcutane-

ous tissues. This may partly explain the increase in the

differences between the TcB readings and TSB

values at higher TSB levels. It is therefore important that

due caution is exercised not to use TcB readings alone in

decision-making at TcB readings above 15mg/dl. More

studies are required in this respect. The GA of the pre-

term neonates in the present study did not significantly

affect the relationship between TcB readings and TSB

levels. The difference in the correlation coefficients be-

tween the various GA categories of preterm neonates (r

Table 2: Table showing the comparative statistics of TSB and

TcB for different birth weight categories.

Birth

weight