Understanding neonatal phototherapy

Neonatal jaundice can affect up to 60% of full term babies during their first week of life. In the UK, babies are routinely monitored following birth for signs of jaundice. A prominent feature in jaundiced babies is the yellowing of the sclera caused by excess bilirubin, and where it is the most superficial part of the sclera  with a  higher elastin content that is considered to more highly absorb bilirubin.  

Interest has been shown¹⁹ in using the degree of yellow pigmentation of the sclera as a measure of the level of bilirubin present in the neonate where there are advantages relating to optical measurements based on the lack of melanin in scleral tissue and also its relatively low level of vascularisation.  

Ongoing developments of such detection techniques using smartphones²⁰  have  focused particularly in low resource settings where access to conventional assay techniques  for bilirubin may be limited. 

As the organ of light detection, the eye continues to reveal the complexities of its multilevel cellular interfaces that enable vision. The interaction of optical radiation with the body as a whole is perhaps one of greater complexity, based on the much more diverse cell types involved.  

This complexity certainly extends from infrared radiation to the ultraviolet (UV) spectrum, where the increasing short wavelengths of UV radiation in sunlight can disrupt DNA within skin cells. The notion that sunlight is beneficial to health can readily be traced back to Greek medicine.  

Typically, the most preferred holiday destinations offer the attraction of an abundance of sunshine. Even today, however, the actual effect of exposure of the body to light is incompletely understood.  

One of the early pioneers of using light as a clinical therapy was Niels Ryberg Finsen,1 founder of modern phototherapy and where UV light was used to treat the condition of lupus vulgaris, a skin condition caused by Mycobacterium tuberculosis. This was a distressing, disfiguring condition and the success of the treatment was greatly appreciated, with numerous treatment centres being established in Europe and the US from 1896 onwards. For his efforts Finsen was awarded the Nobel Prize for Medicine in 1903. 

Rochford General Hospital 

While UV light would be further developed to treat a range of skin conditions such as psoriasis, there was no significant therapy role identified of optical radiation within the visible spectrum. This was to change, however, during the summer months of 1956. Richard Cremer, at this time, was a registrar in charge of the premature baby unit at Rochford General Hospital in Essex where there was access to an open courtyard.  

The local sister, Jane Ward, was mindful to expose her charges on occasion to fresh air and sunshine in this environment, whereupon it was noticed during a ward round that sunshine could change the skin appearance of a baby with the condition of neonatal jaundice. It was commented that an infant was ‘pale yellow except for a strongly demarcated triangle of skin very much yellower than the rest of the body.’ The triangle of yellow on the skin had been an area not exposed to sunshine.  

At the time, this observation was not taken seriously. A more significant observation, however, was that the bilirubin level in a sample of neonatal blood with a known high level of bilirubin had been inadvertently exposed to sunshine and found to have a surprisingly low value. Preliminary observations indicated that exposure of babies with neonatal jaundice to periods of sunshine reduced the bilirubin values.  

Subsequent observations in the local pathology laboratory confirmed the light absorption profile of bilirubin, with a classic peak at around 450nm. A type of fluorescent tube was identified as an artificial source of a suitable light spectrum and a prototype lamp unit was constructed by the local hospital engineers and suspended over a neonate.  

A paper describing the beneficial effect of the use of the lamp unit in reducing bilirubin levels in neonates was subsequently published in The Lancet in 1958,² though it would be another 10 years before the concept found its way into routine clinical practice.  

It is interesting, however, to reflect on the more ‘informal’ approach to hospital practice almost some 70 years ago. In today’s health and safety framework, babies would not be wheeled out into an open courtyard and hospital engineers would not be allowed to make a novel lamp involved in clinical therapy. Today the contribution of Sister Jean Ward resonates³ with those with a memory of earlier times.  

The Haemoglobin Cascade 

A basic question arises, why are bilirubin levels in the neonate important? The answer being that elevated levels can result in irreversible neurological damage as free bilirubin molecules cross the blood brain barrier and damage neural pathways.^{4, 5} In the worst presentations, the condition can be fatal.  

After birth, the neonate changes from a blood supply from the mother to its own respiratory circuit and in this transition, there is a reduction in the number of red blood cells required for oxygen transport. Figure 2 (right) summarises the process involved in the production of bilirubin from the breakdown of haemoglobin.  

Bilirubin is a relatively small molecule and is mostly attached to much larger albumin molecules in the blood. A natural pathway for elimination of albumin bound bilirubin takes place in the liver where it undergoes a transformation to ‘conjugated’ or ‘direct’ bilirubin and with elimination through the gut via the bile duct as indicated in figure 3 (left).  

However, this mechanism is typically not sufficiently effective to deal with the high rate of breakdown of haemoglobin, with the result that serum bilirubin levels have the potential to rise to elevated levels in the days after birth.  

While excretion via the bile duct is considered the main route for elimination of conjugated bilirubin, there is also identification⁶ of the existence of a renal pathway, which could provide additional diagnostic information in cases of neonatal  hyperbilirubinaemia. 

It has been identified^{7, 8} that inherited factors such as Gilbert-Meulengracht, Crigker-Najjar, Dublin-Johnson and Rotor syndrome can affect metabolic pathways that influence the mechanisms of removal of ‘mixed’ (unconjugated and conjugated) and separately unconjugated bilirubin. This identifies elements of potential variability of metabolic function that drive underlying variations in presentations of hyperbilirubinaemia in neonates.  

While it is the ‘free’ bilirubin levels that present the direct clinical risk, bilirubin levels are referenced by the total serum bilirubin value of the ‘free’ and that component bound to albumin. Frequent sampling of bilirubin can be via capillary samples of small volumes of around 35 microlitres. The minimal level of sampling is important since the neonate is not able to produce new blood cells until three or four weeks after birth.  

Every national system of medicine has an adopted set of treatment nomograms for neonates of specific gestational age, where the available treatment options are phototherapy or exchange transfusion and where the use of neonatal phototherapy will usually provide an effective treatment pathway. In the United Kingdom, this is provided by guidance from NICE.⁹  

In general terms, the earlier the term of delivery, the greater the level of intervention required to reduce bilirubin levels. Figure 4 indicates a representative track of bilirubin levels as a function of time for a neonate where the rising level of bilirubin is moderated by an initial phase of phototherapy followed by a subsequent episode, after which the bilirubin level is considered to be in a stable condition with no further requirement for treatment.  

In figure 4, the zone above the red dashed line indicates the requirement to undertake an exchange transfusion, while that zone between the red and black dashed lines indicates the requirement to undertake phototherapy.  

Figure 4: Identification of likely variation of bilirubin (blue line) during neonatal phototherapy

Interactions with Light 

The interaction of bilirubin with light is highly complex. A key element, however, of the process is the formation of isomers of bilirubin, which are molecules of bilirubin in altered energy states based following interaction with light photons. These isomers can be both less toxic and provide an additional pathway for eventual elimination of bilirubin from the neonate’s system.  

While the peak of light absorption of bilirubin-albumin in vitro has been demonstrated at around 459nm, work by Lamola et al, identified that the light absorption in vivo is dominated by the much higher level by haemoglobin, which shifts the peak of absorption to around 476nm.¹¹  

This model was later confirmed¹² by observing fluorescence spectroscopy of bilirubin loaded blood in vitro. Manufacturers of neonatal phototherapy equipment appear, however, reluctant to move the peak irradiance of phototherapy lamp systems to this value.  

Measurement Issues 

While some health clinicians have developed a keen interest in developing neonatal phototherapy, this has not included a focus on spectral measurement parameters of the associated lamp systems. In part this has contributing factors from the device manufacturers and specific device standards.¹² 

Some focus on clinical management and related measurement parameters has been provided by the American Academy of Pediatrics (AAP),¹³ which identified the spectral range between 460nm and 490nm as being clinically relevant with an average value of within this wavelength range expressed as a minimum therapy level as indicated in figure 5. A revision of this guidance has been recently issued¹⁴ where the peak wavelength of response is now identified at 478nm.  

Figure 5: Spectral measurements involving AAP 2011 definition

Meters provided by manufacturers to check the output levels of their devices typically had no common standard of performance or accuracy, so that a specific lamp output would record different values on different meters.¹⁵ Various clinical studies would compare neonatal therapy systems using inappropriate output meters without traceable calibrations to national or international standards.  

The proper approach to such issues would be to measure the spectral output of lamp systems using a spectroradiometer that could, for example, make measurements of irradiance at single wavelength intervals.

A key parameter of such spectroradiometer devices is the spectral resolution based on the ability to resolve contributions at specific wavelengths. It is then possible to determine the spectral output within a selected range of wavelengths. With developing technology, compact hand held spectroradiometers¹⁶ have now become available.  

While the spectral resolution of hand held units is poorer than laboratory based systems, it is still adequate for such clinical applications. The CSS-45-Bili spectroradiometer device indicated in figure 6 is a general purpose handheld spectroradiometer with a specific derived parameter of the average irradiance in the 460nm to 490nm wavelength range in accordance with AAP guidelines.

Figure 7 indicates the screen mode when the device is interfaced to a PC, to allow review and storage of spectral data. In an interpretation of the way in which the output parameter is calculated, is that it is the sum of irradiance values between 461nm and 489nm with inclusion of irradiation in wavelength range 460nm to 460.5nm and the component 489.5nm to 490nm. The irradiance value at 465nm, for example, is interpreted as the irradiance between 464.5nm and 465.5nm.  

Defining Dosimetry 

A basic feature of more formal disciplines of phototherapy is to be able to determine a level of dose of delivered light. Thus in treatment of, for example, psoriasis, there will be a level of dose expressed in joules per square centimetre for a specific UVA or UVB lamp type. For this form of treatment, this concept of local delivery dose is appropriate, since the incident light has its therapeutic effect locally within the superficial skin layers.  

In the case, however, of neonatal phototherapy, the delivered light dose is considered to act upon the intra vascular and the extra vascular component of bilirubin distributed throughout the body of the neonate. While the zone of immediate light interaction is in the first few millimetres of the skin, perfusion of blood within this volume is eventually communicated throughout all the tissues of the neonate.  

A relevant parameter is therefore the rate of light energy incident over the entire surface area of the neonate and not just the reading of a meter held over the neonate. It is estimation of such parameters within specific wavelength bands that can provide objective evaluation of the output of neonatal phototherapy lamp systems.  

During a course of neonatal phototherapy, this would allow a track of the level of reduction of bilirubin level for a given level of delivered energy over the surface of the neonate. This would allow indication of the relative level of response and could identify aspects of metabolism that could be hindering elimination of excess bilirubin and would assist in providing the appropriate level of phototherapy.  

Dosimetry Developments 

Various developments have been undertaken to provide indications of lamp outputs where in the first instance a prototype system was developed, which used a series of 192 photodiodes distributed over the surface of a term manikin¹⁷ where the light dose was determined over 12 specific anatomical areas of a term manikin surface as indicated in figure 8. 

Figure 8: Original ‘babyshape’ with utilisation of 192 individual silicon photodiodes

In order to provide more accurate measurements over curved surfaces, the same ‘babyshape’ was covered with flexible solar film¹⁸ as indicated in figure 9. This provided a more convenient measurement system and where in both systems the system was calibrated using a spectroradiometer. 

In addition, the measurement process used with solar film elements is simplified due to the reduced number of active channels (18 compared with 192). A key element of the simplified construction of the device using solar films was that the detected parameter of generated current could be accurately summed from the separate detection elements connected in parallel. Groups of flexible solar cell elements in a given anatomical area are of similar type and selected to have closely matching spectral sensitivity profiles. Figure 9: ‘Babyshape’ of neonatal term manikin with flexible amorphous silicon elements and also ‘calibration pod’:

One observation of such measurements of the delivered power of light over specific anatomical areas is how the average power values vary over the surface anatomy. For overhead lamp sources, higher values are observed over surfaces as such as the upper torso region compared with the upper and lower leg areas, since the underside of the leg is shielded from the overhead lamp system.  

This indicates that while higher intensity lamps would provide an improvement, the effect of providing a 360 degree delivery mode would also bring benefits, from, for example, wrap round delivery systems.  

Perfusion and Reflection 

It is also relevant to consider the mechanism whereby the isomers of bilirubin become distributed throughout the neonate in the vascular and intravascular compartments, where a basic assumption would be that the rate of conversion of bilirubin is directly proportional to its level of concentration of bilirubin in the surface layers of perfused blood.  

This implies that as the levels of normal bilirubin are ‘bleached’ in this superficial layer, then a key parameter becomes the rate of perfusion of blood between the peripheral layers and the body reservoir of bilirubin.  

This can in effect become the limiting factor of the rate of bilirubin reduction. There is an indication that more effective treatment relates to treating as large a surface area as possible, and not just limited areas at high intensity.  

There is also consideration of reflection of light from the surface of the neonate’s skin surface where as much as 30% of the incident light from phototherapy lamps will be effectively reflected back without therapy effect. This level of reflection is reduced with higher levels of skin pigmentation, though this does not imply increased levels of light involved in conversion of bilirubin.  

A component of such reflected light is in fact light that is scattered in the surface layers of the skin and re-emerges without effective tissue interaction. This also identifies the advantage of ‘wrap round’ products that could include a reflective layer to reflect light back onto the surface of the neonate.  

Summary 

There is an unrivalled array of modern technology now available to provide innovation in the quest for more effective neonatal phototherapy. Underlying such developments, however, is a basic requirement to provide a provide a valid measurement framework to quantify levels of ‘all over’ dose rates provided during neonatal phototherapy. There is also relevance to improve treatment efficiency of existing systems by improving the basic levels of understanding of how to deliver light more effectively to the neonate’s surface.  

References 

1. Grzybowski A, Pietrzak K. From patient to discoverer – Niels Ryberg Finsen (1860–1904) – the founder of phototherapy in dermatology. Clin Dermatol. 2012 Jul-Aug;30(4):451-5  
2. Cremer RJ, Perryman PW, Richards DH. Influence of light on the hyperbilirubinaemia of infants. Lancet 271(7030):1094-1097,1958
3. Maisels MJ. Sister Jean Ward, phototherapy, and jaundice: a unique human and photochemical interaction. J Perinatol. 2015 Sep;35(9):671-5. 
4. Cashore WJ. The neurotoxicity of bilirubin. Clin Perinatol. 1990;17: 437-47
5. Shapiro SM, Riordan SM. Review of bilirubin neurotoxicity II: preventing and treating acute bilirubin encephalopathy and kernicterus spectrum disorders. Pediatr Res. 2020; 87: 332-337
6. Thomas M, Hardikar W, Greaves RF, Tingay DG, Loh TP, Ignjatovic V et al, Mechanism of bilirubin elimination in urine: insights and prospects for neonatal jaundice. Clin Chem Lab Med. 2021 Jan 15;59(6):1025-1033 
7. Strassburg CP. Hyperbilirubinemia syndromes (Gilbert-Meulengracht, Crigler-Najjar, Dubin-Johnson, and Rotor syndrome). Best Pract Res Clin Gastroenterol. 2010 Oct;24(5):555-71. 
8. Watchko JF, Tiribelli C. Bilirubin-induced neurologic damage--mechanisms and management approaches. N Engl J Med. 2013 Nov 21;369(21):2021-30
9. NICE, Jaundice in newborn babies under 28 days Clinical guideline [CG98]Published: 19 May 2010 Last updated: 31 October 2023 : https://www.nice.org.uk/guidance/cg98/resources 2023 [accessed 13th January 2025]
10. Lamola AA, Bhutani VK, Wong RJ, Stevenson DK, McDonagh AF. The effect of hematocrit on the efficacy of phototherapy for neonatal jaundice. Pediatr Res. 2013 Jul;74(1):54-60
11. Lamola AA, Russo M. Fluorescence excitation spectrum of bilirubin in blood: a model for the action spectrum for phototherapy of neonatal jaundice. Photochem Photobiol. 2014 Mar-Apr;90(2):294-6
12. International Electrotechnical Commission, Medical electrical equipment – Part 2-50: Particular Requirements For The Basic Safety And Essential Performance of Infant Phototherapy Equipment, 2023, Geneva.
13. Bhutani VK; Committee on Fetus and Newborn; American Academy of Pediatrics. Phototherapy to prevent severe neonatal hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics. 2011 Oct;128(4):e1046-52.
14. Kemper AR, Newman TB, Slaughter JL, Maisels MJ, Watchko JF, Downs SM et al, Clinical Practice Guideline Revision: Management of Hyperbilirubinemia in the Newborn Infant 35 or More Weeks of Gestation. Pediatrics. 2022 Sep 1;150(3):e2022058859. 
15. Clarkson DM, Nicol R, Chapman P. Neonatal phototherapy radiometers: current performance characteristics and future requirements. Med Eng Phys. 2014 Apr;36(4):522-9
16. Clarkson DM, Satodia P. Use of a hand-held spectroradiometer for the measurement of neonatal phototherapy lamp outputs. Med Eng Phys. 2019 Nov;73:107-111.
17. Clarkson DM, Tshangini M, Satodia P. Preliminary observations of a system for determination of phototherapy exposure over a neonate body shape. Med Eng Phys. 2021 Sep;95:1-8.
18. Clarkson DM and Satodia P. The use of solar film elements on a neonate manikin surface to estimate the received output power of neonatal phototherapy lamp systems – IPEM-Translation 2024, Volumes 10–11;100029. 
19. Leung TS, Kapur K, Guilliam A, Okell J, Lim B, MacDonald LW, Meek J. Screening neonatal jaundice based on the sclera color of the eye using digital photography. Biomed Opt Express. 2015 Oct 23;6(11):4529-38
20. Outlaw F, Nixon M, Odeyemi O, MacDonald LW, Meek J, Leung TS. Smartphone screening for neonatal jaundice via ambient-subtracted sclera chromaticity. PLoS One. 2020 Mar 2;15(3):e0216970.