Transcutaneous bilirubinometry during and after phototherapy

KL Tan1 and F Dong2

Specialist Infant–Child Centre1, Mt. Elizabeth Medical Centre; Department of Biostatistics2, National University Medical Institute,Singapore

Tan KL, Dong F. Transcutaneous bilirubinometry during and after phototherapy. Acta Pædiatr2003; 92: 327–331. Stockholm. ISSN 0803-5253

Aim: To evaluate the reliability of transcutaneous bilirubinometry (TcB) during and after photo-therapy. Methods: TcB was performed on the forehead and chest of infants with neonatal jaundicewhen capillary blood was sampled for bilirubin determination in a control group of 240 neonates.In a second group of 70 neonates exposed to phototherapy the same procedure was performed afterat least 24 h of exposure; on the forehead, TcB was done in the centre of the unexposed area andalso on the adjacent exposed area, and the exposed chest. During the post-phototherapy period,TcB was again done during the first and second days, at least 18–24 h after cessation of photo-therapy. The results were then statistically evaluated and regression curves were plotted. Results:A close correlation between TcB values and bilirubin levels was observed in the control group. Inthe phototherapy group, a correlation was also found between the TcB and the bilirubin values, butthis correlation was significantly poorer than that of the controls; the correlation for the coveredpart of the forehead was significantly better than that of the exposed part but still poorer than thatof the controls, though the difference was no longer significant. Skin colour recovered during thepost-phototherapy period and correlation was better than that during exposure and no longersignificantly different from that of the controls.

Conclusion: Through its bleaching effect on the skin, phototherapy affects the correlationbetween TcB and the bilirubin values, but does not totally eliminate it. The unexposed parts showa better correlation, though this was still poorer than that of the controls. Recovery of skin colouroccurred within 18–24 h after cessation of exposure.

Key words: Transcutaneous bilirubinometry, phototherapy

KL Tan, Specialist Infant–Child Centre, 3 Mt Elizabeth, 17–16 Mt Elizabeth Medical Centre,228510 Singapore (Tel. �65 735 9544, fax. �65 733 2551, e-mail. ylkltan@singnet.com.sg)

Neonatal jaundice is a common complication ofneonates, especially among the Asian races (1–3); theevaluation of its severity is therefore an importantaspect in its management. Transcutaneous bilirubin-ometry (TcB) is an easy, safe and convenient methodfor the evaluation of the severity of jaundice, since noinvasive procedure is involved. Several studies havedocumented its reliability in this respect (4–7), even ininfants with very dark complexions (7). Though inmany studies it is claimed that TcB is affected byconventional phototherapy (5, 6, 8–10), no evaluationof TcB during and after exposure to phototherapy has asyet been made. The purpose of this study was toevaluate the impact of phototherapy on the overallaccuracy of TcB during and after phototherapy usingnon-exposed infants as controls.

Materials and methodsFull-term, healthy mature infants admitted to the lying-

in nursery, with clinical evidence of jaundice (with notrauma or other complications), were enrolled in thestudy. TcB was performed using the Minolta-Airshieldsbilirubin meter, which was checked daily against thestandard control before being used. This TcB procedurewas performed in the morning under the normal nurserylighting conditions, away from any strong source oflight, when capillary blood sampling was performed.TcB was done on the forehead (glabella) and over thesternum. Three readings were taken for each site; onlywhen the values did not deviate by more than 10% werethe readings recorded, and the mean value taken as thefinal reading. Where repeat bilirubin determinationswere required on successive days because of risingbilirubin concentrations, the same procedure wasfollowed. It was deemed that since the procedures wereperformed at least 24 h apart, the different readingscould be accepted as distinct entities independent of oneanother, as the situation and time-frame had changed;this was therefore very similar to the situation wherebilirubin concentrations are determined for different

2003 Taylor & Francis. ISSN 0803-5253

���� ������ �� ��� �� ���

infants (7). The number of TcB and bilirubin determi-nations performed was therefore greater than thenumber of infants studied.

When bilirubin values were �255 �mol/L or�222 �mol/L in the first 72 h of life, phototherapywas commenced. The phototherapy setup consisted ofseven daylight lamps (Philips 20W/54) arranged in anarc, 35 cm above the infants’ skin; the radiance at skinlevel was 250 microwatts/cm2 (�W/cm2) in the 425–475 nanometer (nm) range, and 295 �W/cm2 in the400–480 nm range and total light intensity 4630 lux atskin level, as previously described (11). During photo-therapy, the infants’ eyes were routinely covered byopaque eye pads to shield them from direct exposure tothe phototherapy light. The pads were 6 cm wide and17 cm long and attached by micropore tape to the side ofthe head. The nursing personnel routinely checked theposition of the eye pads at 3-h intervals during feeds.Before the TcB procedure and capillary bilirubinsampling were performed, the phototherapy lights wereswitched off, and the infants removed from the cots. Theforehead TcB determination was performed on thecentre of the covered glabella area as well as on theadjacent exposed area, while the chest TcB wasperformed on the sternum (in the same manner as inthe controls), which was completely exposed duringphototherapy. During phototherapy exposure, TcBdetermination was performed only after a minimumexposure period of 24 h. Repeat TcB determinationswere performed only after an interval of at least 24 h.Again, it was felt that after such a long interval, thedeterminations could be assumed to be different inde-pendent entities. After cessation of exposure, the infantswere observed for two more days before being dis-charged, in order to monitor the rebound values; duringthis period the TcB was again performed when thecapillary sampling was done; the time interval betweencessation of phototherapy and the first post-photother-apy TcB varied between 18 and 24 h, and that betweenthe first and second post-phototherapy TcB was 24 h. Inthis way the length of time required for the skin toreverse the bleaching effect of the exposure, and thusagain reliably reflect the degree of jaundice, could alsobe evaluated. To evaluate the reliability and reproduci-bility of the TcB jaundice meter, the coefficient ofvariation was evaluated. Ten consecutive readings at thesame site were obtained for 5 different infants.

To minimize confounding variables, only healthyfull-term Chinese infants were enrolled in the study.Informed consent was obtained.

The capillary blood samples were placed in red-coloured straws and kept in a light, tight box until themoment of determination; this was done to prevent anybleaching of the blood by environmental light. Plasmabilirubin determination was performed under controlledconditions using the AO Bilirubinometer (AmericanOptical, USA), which was calibrated regularly againstknown controls.

The results were evaluated statistically. The Pearsoncorrelation coefficient was utilized to determine thestrength of any association between the variables. Theassessment of the equality of the gradient and interceptsof the linear regression curves, as well as the correlationvalues, was determined by means of a multipleregression analysis.

The Ethics Committee of the Singapore NationalUniversity Hospital approved the study.

ResultsAll the infants enrolled were healthy before, during andafter the study and they were all full-term infants withpostnatal ages varying between 2 and 7 d (see Table 1).There were no complications associated with theprocedures. The reliability of the Minolta-Airshieldsbilirubin meter, as evaluated by 10 consecutive readingsover the same site on any one occasion, demonstratedsimilar values in five infants with a coefficient ofvariation of only 5%. A significant correlation wasfound between the TcB and the bilirubin values in thecontrol group not exposed to phototherapy, (Fig. 1). Thespread of the values was very close to the regressionslope, with none widely dispersed.

In the group exposed to phototherapy, there wasobvious bleaching of the skin, to the extent that visualevaluation of the degree of jaundice was virtuallyimpossible. The TcB readings on the exposed areas forthe different plasma bilirubin values were highlyreproducible for the different durations of exposure,the minimum being 24 h. Yet in this situation, asignificant correlation between the TcB readings andthe plasma bilirubin values (Fig. 2) was observed. Butthe correlation coefficient was significantly less thanthat of the control group (chest, r = 0.60 vs r = 0.88,p � 0.001; forehead (exposed), r = 0.60 vs r = 0.80,p � 0.001). The correlation between the TcB of thecovered part of the forehead and the bilirubin values(r = 0.74) was more marked than that of the exposedpart (p � 0.05), and though less than that of the controlgroup, this was not significantly different (0.16 � p �0.15). This covered area was definitely darker than theexposed area but a distinct demarcation line between theexposed and covered parts of the forehead could not beobserved. The slopes of the correlation curves of theexposed parts of the chest and forehead were notsignificantly different from those of the control curves.

Table 1. Clinical data of infants studied.

Birthweight (g)Gestation

(wk) Age (d)No. (M:F) Mean � SD Mean � SD Mean � SD

Controls 240 (131:119) 3237.7 � 425.0 39.1 � 1.3 4.1 � 1.1Photo 70 (37:33) 3252.2 � 428.8 39.9 � 1.2 5.1 � 1.2

328 KL Tan and F Dong ACTA PÆDIATR 92 (2003)

Fig. 1. Correlation between TcB_C (chest) and TcB_F (forehead) and bilirubin values in controls.

Fig. 2 . Correlation between TcB_NE (non-exposed part forehead)and TcB_E (exposed part forehead) and TcB_C and bilirubin valuesduring phototherapy.

ACTA PÆDIATR 92 (2003) TcB during and after phototherapy 329

However, the intercepts were significantly differentbetween the control and exposed groups for the chestmeasurements (p � 0.001) and the exposed part of theforehead (p � 0.002), while that for the non-exposedpart of the forehead was not significantly different.

In the post-phototherapy period, there was a rapidreturn of the jaundice colour. The covered and exposedareas appeared similar in colour, resulting in improvedcorrelation on day 1 (chest r = 0.80; forehead, r = 0.70),but the values were no longer significantly differentfrom those of the control group (0.07 � p � 0.06 and0.13 � p � 0.12, respectively), and further improve-ment on day 2 (chest r = 0.84, 0.40 � p � 0.30; fore-head, r = 0.80, p = 1.00). However, the differencesbetween the values on day 1 and those on day 2 werenot significantly different. There was no difference inthe TcB readings between the exposed and unexposedparts of the forehead from the first day after photo-therapy. In fact, the skin colour between the adjacentareas was indistinguishable. The curves for the first andsecond days were similar in both the spread of thevalues and the correlation coefficients; the two sets ofvalues were therefore combined and evaluated together(Fig. 3). No significant difference was found betweenthe pre- and post-phototherapy curves in the correlationcoefficients, slope or intercepts.

DiscussionIn the control group, the close correlation betweenplasma bilirubin and TcB is again a further confirmationof earlier studies showing TcB accurately to reflect theplasma bilirubin concentration in the full-term infant.Both the readings on the forehead and sternum demon-strated a good correlation. It can thus be used as a fairly

reliable screening method for detecting significantjaundice that might need further evaluation.

With phototherapy, such a correlation still exists butto a significantly lesser extent; correlation was sig-nificantly poorer, in line with observations elsewhere(5, 6, 8). The usefulness of phototherapy in a clinicalsituation is limited because of the wide dispersal of thevalues. The correlation observed in the area of theglabella not exposed to phototherapy showed a betterrelationship than that of the exposed part, with acorrelation comparable to that in the control group,and with the points more closely aggregated. Since thepad was not glued on, only at each end, it was highlylikely that some “seepage” of light along the edgeswould occur, leading to some exposure during photo-therapy. Hence no definite demarcation margin could beobserved between the covered and exposed parts,though the covered area was obviously more jaundiced.This, however, should not have affected the TcB valuesthat were taken in the centre of the covered part.Furthermore, the possibility of the pad slipping a littleduring exposure should not have affected the centralreading, as the pad was fairly wide. Commercial eyepads are available, but light seepage and slipping of thepads are still possible. The slight decrease in correlationof the unexposed part compared with that in the controlgroup is an indication of the equilibration process beingdelayed with regard to the skin, where the dermalbinding of the bilirubin appears to be “firmer”. Inprevious reports it is claimed that there was little changein the unexposed areas of the skin during the first twohours of exposure (9), but this became noticeable atlonger time intervals (9, 10, 12). However, in thepresent study we demonstrated that the correlationcoefficient of the unexposed parts of the skin duringphototherapy, though slightly less, was still not sig-

Fig. 3. Correlation between TcB_C and TcB_F and bilirubin values post-phototherapy.

330 KL Tan and F Dong ACTA PÆDIATR 92 (2003)

nificantly different even after 24 or 48 h of phototherapyfrom that of the controls. The unexposed parts of theskin still reflect the degree of jaundice fairly well,though not as markedly as in the controls.

With cessation of exposure, a return of the jaundicecolour to the skin improved the degree of correlation somarkedly that by 18–24 h, the correlation coefficientwas no longer significantly different from that of thecontrol group. However, a further slight improvementoccurred on the second day resulting in the correlationbeing highly in comparison with that of the controls.Difference in colour between the covered and exposedarea of the forehead could no longer be observed by thefirst post-phototherapy day. Obviously, one or (evenbetter) two days were required for a complete return ofskin coloration to the previous degree of reliability. Thisis an indication of the process of equilibration in bili-rubin dynamics, which in the skin took a longer timeinterval than in the plasma. This is further confirmationof the lag in dermal bilirubin deposition and depletioncompared with in plasma. Indeed this explains thedelayed though still marked efficacy of phototherapy indecreasing bilirubin values.

The intercepts in the exposed areas of the photo-therapy group were significantly different from those ofthe control group, with, again, the unexposed areasbeing not significantly different from the controls. Thebleaching effect of the exposure had decreased theintercept significantly. However, with cessation ofexposure, the re-pigmentation of the skin was sufficientto ensure comparability to that of the controls.

Though pre-phototherapy, the correlation betweenTcB and plasma bilirubin values was good, thecorrelation during exposure was relatively weak, andthe points widely dispersed. The usefulness of TcBduring phototherapy for determining whether exposurecan be terminated is therefore likely to be limited. Thereis a stronger case in the controls for the use of TcB todetermine the need for phototherapy, though confirma-tion by laboratory determination would provide greaterreassurance. Furthermore, since inter-ethnic variationsdo occur, curves for different ethnic groups are needed.

A full evaluation of the effect of phototherapy on skincolour has as yet not been reported. The present studydemonstrates that though bleaching of the exposed skinoccurs during phototherapy, its impact does not totallyinvalidate the relationship between the exposed skin andthe degree of jaundice of the infant. Binding of bilirubin

to the skin is thus quite firm and not readily completelyremoved even with continuous exposure for more than24 h. However, the covered skin is almost as good as theskin of the controls as an indicator of jaundice. Theyellow restaining of the skin is apparently gradual, asdemonstrated by the post-phototherapy observations:18–24 h before almost complete restoration of skincolour occurred, a rate that would still enable the skin tobe a good indicator of the degree of jaundice even incases with rapid onset, as long as exposure to verybrightly lit conditions does not occur. During photo-therapy, TcB of the covered area might still be a fairlyreliable indicator of the degree of jaundice.

Acknowledgements.—We thank the nurses in the wards for theirenthusiastic cooperation in this study, and the Ethics Committee of theNational University Hospital for its encouraging approval of this study.

References1. Tan KL, Glucose-6-phosphate dehydrogenase status and neo-

natal jaundice. Arch Dis Child 1981; 56: 574–72. Brown W, Wong HB. Ethnic group differences in plasma bili-

rubin levels of full-term healthy Singapore newborns. Pedatrics1965; 36: 745–9

3. Osborn LM, Reiff MI, Bolus R. Jaundice in the full-term neonate.Pediatrics 1984; 73: 520–5

4. Yamanouchi I, Yamauchi Y, Igarashi I. Transcutaneous bili-rubinometry: preliminary studies of non-invasive transcutaneousbilirubin meter in the Okayama National Hospital. Pediatrics1980; 65: 195–212

5. Tan KL. Transcutaneous bilirubinometry in Chinese and Malayinfants. Acta Paediatr Scand 1982; 71: 593–6

6. Heygi T, Hiatt IM, Indyk L. Transcutaneous bilirubinometry. 1.Correlation in term infants. J Pediatr 1981; 98: 454–7

7. Tan KL, Chia HP, Koh BC. Transcutaneous bilirubinometry inChinese Malay and Indian Infants. Acta Pædiatr 1996; 85: 986–9

8. Tudehope DI, Chang A. Non-invasive method of measuringbilirubin levels in newborn infants. Med J Aust 1982; 1: 165–7

9. Cifuentes RF, Nelson AJ, Levine J, et al. Cutaneous bilirubin-ometry during phototherapy. Pediatr Res 1982; 16: 282A

10. Pasnick M, Luce JF. Transcutaneous bilirubinometry can be usedduring phototherapy. Pediatr Res 1982; 16: 32A

11. Tan KL. Comparison of the effectiveness of single direction anddouble direction phototherapy for neonatal jaundice. Pediatrics1975; 56: 550–3

12. Tolentino T, Zanni R Gertner I, et al. Transcutaneous bilirubin-ometry: correlations during phototherapy. Pediatr Res 1982; 16:312A

Received July 2, 2002; revision received Sept. 23, 2002; acceptedOct. 4, 2002

ACTA PÆDIATR 92 (2003) TcB during and after phototherapy 331