Unconjugated hyperbilirubinemia in preterm infants <35 weeks gestation

INTRODUCTION — Almost all preterm infants less than 35 weeks gestational age (GA) have elevated total serum/plasma bilirubin (TB) levels. When adjusted for gestational and postnatal ages, preterm infants are inherently at greater risk than more mature infants (those born term or late preterm) for developing bilirubin-induced neurologic dysfunction (BIND), which, if not treated in a timely or appropriate manner, can result in chronic neurologic sequelae.

Hyperbilirubinemia in the preterm infant <35 weeks GA will be reviewed here. The screening, evaluation, and management of hyperbilirubinemia in the late preterm and term infant are discussed separately. (See "Unconjugated hyperbilirubinemia in neonates: Risk factors, clinical manifestations, and neurologic complications" and "Unconjugated hyperbilirubinemia in term and late preterm newborns: Screening" and "Unconjugated hyperbilirubinemia in term and late preterm newborns: Initial management".)

BILIRUBIN-INDUCED NEUROLOGIC DYSFUNCTION — Clinical observational data in neonates have reported a spectrum of neurologic conditions among vulnerable neonates who have experienced an exposure to bilirubin of a lesser degree than generally associated with acute bilirubin encephalopathy (ABE) and chronic bilirubin encephalopathy (CBE), respectively. This syndrome of bilirubin-induced neurologic dysfunction (BIND) is a major complication of an elevated total serum/plasma bilirubin (TB) level. Subtle clinical neuromotor manifestations include processing disorders with objective perturbations of visuomotor, auditory, speech, cognition, and language. Risk factors include prematurity, presence of hemolysis, perinatal-neonatal complications, altered bilirubin-albumin binding, and severity and duration of bilirubin exposure.

BIND occurs when unconjugated bilirubin, which is not bound to albumin (also referred to as "free" or "unbound bilirubin" [UB]), crosses the blood-brain barrier, enters the brain, and causes neurologic injury (figure 1). Neuronal damage affects several areas of the brain, including the pontine brainstem oculomotor nuclei globus pallidus, auditory pathways, hippocampus, diencephalon, subthalamic nuclei, cerebellum, and the vermis. The ensuing damage can result in acute bilirubin encephalopathy, and, if not treated, results in the posticteric neurologic manifestations that present as a chronic bilirubin encephalopathy (CBE), previously referred to as kernicterus [1,2]. These include disordered visual gaze (eg, limitations of upward gaze) [3], sensorineural hearing impairment [4], gait abnormalities (choreoathetoid cerebral palsy) [5], and speech and language disorders [6]. Subtle neuromotor signs are associated with a range of processing disorders due to visuo-oculomotor, auditory, speech, and expressive language disturbances. (See "Unconjugated hyperbilirubinemia in neonates: Risk factors, clinical manifestations, and neurologic complications", section on 'Bilirubin-induced neurologic disorders (BIND)'.)

Increased vulnerability due to prematurity — Preterm infants, compared with term infants, are more vulnerable to the syndrome of BIND and kernicterus at lower TB levels [7]. This increased susceptibility was first illustrated by case reports from the 1960s and 1970s in preterm infants with Rhesus isoimmune hemolytic disease [8]. Kernicterus was diagnosed in preterm infants at TB levels ranging from 10 to 18 mg/dL (171 to 308 micromol/L) compared with higher TB values associated with kernicterus in term infants (TB >20 mg/dL [342 micromol/L]).

Limited data demonstrating the increased vulnerability include the following:

●In an analysis using information from several regional and national databases, preterm infants 32 to <37 weeks GA were 20 percent more likely to develop CBE than those born at term [9].

●A large retrospective analysis of extremely low birth weight (ELBW) infants (BW <750 g) reported only a small association between modest levels of TB (5 to 12 mg/dL [85 to 204 micromol/L]) and a combined outcome of death and neurodevelopmental impairment (odds ratio [OR] 1.07, 95% CI 1.03-1.11), hearing impairment, and hearing impairment requiring hearing aids (OR 1.14, 95% CI 1.00-1.30) [10].

●In small case reports of extremely preterm infants, CBE was associated with only moderately elevated TB [7,11-13].

•In one case series, five preterm infants (25 to 29 weeks GA) with peak TB levels ranging from 8.7 to 11.9 mg/dL (149 to 203 micromol/L) had sensorineural hearing loss and magnetic resonance imaging (MRI) abnormalities in the globus pallidus [12].

•In a second report, all nine infants with auditory neuropathy spectrum disorder were born extremely preterm (median: 27.4 weeks GA) and all received phototherapy with a mean peak TB level of 11.9 mg/dL (203.6 micromol/L) [13].

●Indirect evidence from a National Institute of Child Health and Human Development (NICHD) trial in very low birth weight (VLBW) infants (BW <1000 g) showed more aggressive phototherapy (initiated at lower TB thresholds than those for conservative therapy) compared with conservative phototherapy was associated with a reduction of neurodevelopmental impairment (NDI) at 18 to 22 months corrected age [14]. However, a second published subanalysis reported aggressive phototherapy was associated with an increase in mortality for the most preterm infants (ie, ELBW infants [BW<750 g]) who were mechanically ventilated, but the reduction of NDI persisted in survivors [15]. (See 'Phototherapy and mortality' below.)

Hypoalbuminemia and bilirubin binding capacity (BBC) — The increased vulnerability to BIND in preterm infants may be associated with hypoalbuminemia, which results in a decrease in BBC [7,12,16-18]. BBC defines the relationship between an infant's level of "free" or unbound bilirubin (UB) and his/her ability to "tolerate" increasing bilirubin loads. Hypoalbuminemia may be due to decreased synthesis, increased catabolism due to severe illness (eg, sepsis), and abnormal loss [7]. In addition, the decrease in BBC may be related to several factors that negatively affect bilirubin-albumin binding, including asphyxia, acidosis, sepsis, meningitis, and drug therapies that compete with bilirubin-binding sites (eg, sulfa drugs). However, one randomized controlled trial did not find that the use of B/A molar ratios improved neurodevelopmental outcome in infants ≤32 weeks GA [19]. (See 'Other tests' below.)

PATHOPHYSIOLOGY — The pathogenesis of hyperbilirubinemia in the preterm infant is similar to that in the term infant (see "Unconjugated hyperbilirubinemia in neonates: Etiology and pathogenesis"). Hyperbilirubinemia in preterm infants is more prevalent and protracted than that in term infants [8].

It is primarily caused by the following alterations in bilirubin metabolism compared with adults and more mature term infants (figure 1):

●Increased bilirubin production because of increased red blood cell (RBC) breakdown (turnover)

●Decreased bilirubin clearance and conjugation (immature liver)

●Increased enterohepatic circulation of bilirubin

In preterm infants, there also is often a delay in enteral feeds, which may limit intestinal flow and bacterial colonization, resulting in further enhancement of the enterohepatic circulation of bilirubin.

SCREENING FOR HYPERBILIRUBINEMIA

Approach to screening — Screening for hyperbilirubinemia is based on the rationale that early identification and treatment of hyperbilirubinemia will reduce the risk of bilirubin-induced neurologic dysfunction (BIND), especially chronic bilirubin encephalopathy (CBE, previously referred to as kernicterus). In our center, universal screening is performed with total serum/plasma bilirubin (TB) levels in all preterm infants beginning at 24 hours after birth or earlier if the neonate is at increased risk for developing hyperbilirubinemia. Subsequent TB levels are measured every 24 hours while the infant is in the neonatal intensive care unit (NICU) and the frequency is increased to every 12 hours if TB is approaching the threshold level for the initiation of phototherapy (figure 2) [20-22]. For infants <35 weeks GA and not deemed ill, frequency of TB monitoring (by transcutaneous bilirubinometry [TcB] or by clinical lab assays) may be increased based on clinical judgment and risk factors. The clinician should consider using daily monitoring of jaundice progression, periodic TcB testing, and an infant's postnatal age in their assessment in order to minimize overtesting and overuse of phototherapy.

Choice of screening test — Our preferred screening test to identify preterm infants at-risk for hyperbilirubinemia is measuring TB.

Total serum/plasma bilirubin (TB) — TB is the most frequent clinical laboratory test used to measure bilirubin and detect neonatal hyperbilirubinemia and is our preferred choice for screening in preterm infants. Although highly variable, it is the only available surrogate used to predict the risk of CBE (kernicterus). However, in extremely preterm infants, TB may not accurately reflect the concentration of free (unbound) bilirubin (UB) due to reduced and varied bilirubin binding capacity (BBC) for infants [17]. In this setting, UB testing (if available) or measuring the bilirubin/albumin (B/A) molar ratio may be appropriate. However, these tests are not universally available and may be found only in a research setting. (See 'Hypoalbuminemia and bilirubin binding capacity (BBC)' above and 'Other tests' below.)

Transcutaneous bilirubin (TcB) — TcB devices are widely used to screen for hyperbilirubinemia in term and late preterm infants. However, due to limitations in their accuracy and precision, we do not recommend the routine use of TcB devices for assessing hyperbilirubinemia in preterm infants. Although, a 2013 systematic review suggested that TcB devices reported similar reliability in estimating TB in preterm infants less than 37 weeks GA compared with more mature infants [23], subsequent data showed that the correlation between TcB and TB decreases with decreasing gestational age [24,25]. For extremely preterm infants (<30 weeks GA), the correlation of measurements between TcB and TB also varies depending on the body site used due to differences in tissue bilirubin binding [26].

The concern of poorer reliability of TcB with decreasing GA is further heightened because intervention thresholds become narrower with increasing immaturity. As a result, TcB screening is not widely used [27]. Thus, we do not recommend the routine use of TcB devices in extremely preterm infants (gestational age <28 weeks) until there are improved devices with better accuracy and precision, and a clinically validated standardized protocol for its use in preterm infants (GA <35 weeks).

Other tests — Other laboratory tests that are helpful in assessing the risk of hyperbilirubinemia but that are not routinely available include:

●Free or unbound bilirubin (UB) – UB, especially in extremely preterm infants, may be a more appropriate predictor of BIND [17,28-31]. However, laboratory testing of UB is not universally available and is primarily used in a research setting.

●Bilirubin/albumin (B/A) molar ratio – Preterm infants are likely to have low serum albumin levels, and it has been suggested that the B/A molar ratio would be a good measure of the risk of bilirubin toxicity based upon birth weight [32,33]. However, its usefulness has been limited in preterm infants, as other factors (eg, acidosis, use of multiple drugs, elevated free fatty acids, and presence of photoisomers) may interfere with B/A binding or binding of bilirubin to sites other than albumin [30,34]. As a result, we do not routinely check B/A in preterm infants.

CLINICAL PRESENTATIONS

Jaundice — Jaundice is the yellow color produced by the deposition of bilirubin in the skin, subcutaneous tissues, and/or conjunctiva (as visualized on the sclerae). Although an important and time-honored clinical sign, the presence or absence of visible jaundice is not a reliable method to assess total serum/plasma bilirubin (TB) levels (or severity of hyperbilirubinemia). In particular, visual assessment for jaundice cannot be relied upon to identify elevated TB in extremely preterm infants (GA <28 weeks), especially those with dark skin pigmentation, in whom phototherapy is used at much lower TB levels, or infants with rapidly rising TB levels. (See "Unconjugated hyperbilirubinemia in neonates: Risk factors, clinical manifestations, and neurologic complications", section on 'Jaundice'.)

Neurologic findings — Bilirubin-induced neurologic dysfunction (BIND) can present either acutely as acute bilirubin encephalopathy (ABE) or chronically as chronic bilirubin encephalopathy (CBE, previously referred to as kernicterus).

Acute bilirubin encephalopathy (ABE) — In preterm infants, the acute signs (eg, hypertonia, irritability, posturing, arching, and seizures) commonly seen in more mature neonates are not typically observed [35,36]. Preterm infants may only manifest minimal findings of recurrent apnea, periodic breathing, or episodes of oxygen desaturation, which may be masked in mechanically ventilated infants [7]. In one observational study from 1955 of preterm infants with gestational age (GA) >30 weeks, signs and symptoms of ABE observed in the first 24 to 48 hours of birth included head retraction, expressionless facies (usually with oculogyric movements), changes in muscle tone, episodes of cyanosis, and lack of sucking [37].

In preterm infants, there is a disproportionately increased risk for sensorineural hearing loss (SNHL) [38]. As a result, automated auditory brainstem-evoked response (AABR) measurements is used to detect acute reversible bilirubin-induced auditory dysfunction, as well as irreversible SNHL [32]. In our center, an abnormal AABR is often used as another indicator for potential exchange transfusion. We perform AABR in all preterm infants (<36 weeks’ GA) with elevated TB requiring intervention or who were admitted to the neonatal intensive care unit as they are at risk for SNHL, which cannot be detected by otoacoustic emissions. (See "Unconjugated hyperbilirubinemia in neonates: Risk factors, clinical manifestations, and neurologic complications", section on 'Acute bilirubin encephalopathy (ABE)' and "Screening the newborn for hearing loss", section on 'Automated auditory brainstem response'.)

Chronic bilirubin encephalopathy (CBE) or kernicterus — Chronic bilirubin encephalopathy (CBE, previously referred to as kernicterus), is the chronic and irreversible form of BIND with permanent neurologic sequelae. The clinical features of CBE in preterm infants are similar to those observed in term infants, as the globus pallidus remains the primary target of brain injury. In addition to the following characteristic findings of CBE, these patients also display a range of processing disorders with disturbances of visuomotor, auditory, speech, cognition, and language [2]:

●Auditory neuropathy manifested as SNHL (abnormal AABR with normal otoacoustic emissions) – Observational studies suggest that isolated sensorineural hearing abnormalities may be the predominant or the sole manifestation of BIND. (See "Hearing loss in children: Etiology", section on 'Hyperbilirubinemia'.)

●Cerebral palsy (CP) characterized by choreoathetosis. (See "Cerebral palsy: Classification and clinical features".)

●Upward gaze abnormalities.

●Enamel dysplasia of deciduous teeth.

Imaging — In infants with BIND, abnormalities on magnetic resonance imaging (MRI) include increased signaling localized to the dentato-thalamo-cortical pathway [39]. In a small series, changes were seen in the globus pallidus with shifts from T1 increased signaling during the acute phase to permanent T2 hypersignaling and normal T1 signaling at 12 or 22 months corrected age [12].

MANAGEMENT APPROACH

Overview — Therapeutic interventions (ie, phototherapy and exchange transfusion) reduce total serum/plasma bilirubin (TB) levels in the blood and potentially prevent bilirubin-induced neurologic dysfunction (BIND). In term and late preterm infants (gestational age [GA] ≥35 weeks), hour-specific TB nomograms are used to guide decisions regarding initiating phototherapy and performing exchange transfusion [40] (see "Unconjugated hyperbilirubinemia in term and late preterm newborns: Initial management", section on 'Thresholds for treatment').

In preterm infants <35 weeks GA, similar guidelines are not available, despite observational evidence that suggests these preterm infants are more susceptible to BIND at lower TB levels than more mature infants. Nevertheless, based on consensus expert opinion and available data from studies from the National Institute of Child Health and Human Development (NICHD), we initiate therapy, primarily phototherapy, in infants <35 weeks GA and who are <7 days of age using a stratified approach based upon GA and TB [10,14,20,21]. The TB targets are in line with the more aggressive therapy used in the 2008 study based on data suggesting improved neurodevelopmental outcome [10,14].

The lack of an evidence-based consensus on the management of hyperbilirubinemia in the preterm infant is due to variabilities in the clinical manifestations and spectrum of BIND, absence of reliable and predictive measures of bilirubin neurotoxicity, and uncertainties of the relative risks and benefits of interventions to reduce TB levels (ie, phototherapy and exchange transfusions). A variety of approaches, including aggressive (early phototherapy at a low TB threshold level) and conservative treatment, have been proposed and generally are based upon using threshold TB levels stratified by birth weight (BW) [41-45]. (See 'Benefit and risk' below.)

Initiation of phototherapy — We initiate phototherapy for preterm infants based on GA and TB as follows (figure 2) [20,21]:

●GA <28 weeks – TB >5 mg/dL (86 micromol/L)

●GA 28 to 29 weeks – TB 6 to 8 mg/dL (103 to 137 micromol/L)

●GA 30 to 31 weeks – TB 8 to 10 mg/dL (137 to 171 micromol/L)

●GA 32 to 33 weeks – TB 10 to 12 mg/dL (171 to 205 micromol/L)

●GA 34 to <35 weeks – TB 12 to 14 mg/dL (205 to 239 micromol/L)

We suggest using devices with blue light-emitting diodes (LEDs) since they appear to achieve greater reduction in TB compared with other light sources [46]. Phototherapy is initially applied to the ventral body area with uniform exposure (picture 1) at an irradiance range of 15 to 30 microW/cm² per nm, depending on GA.

For rising TB levels, the intensity of phototherapy is increased in incremental steps based on GA as follows:

●>26 to 28 weeks GA:

•Start with 25 to 30 microW/cm²/nm to one (dorsal/ventral) body surface.

•Increase to 25 to 30 microW/cm²/nm to both ventral and dorsal body surfaces.

•Pre-exchange – 30 microW/cm²/nm to both ventral and dorsal body surfaces.

●24 to <26 weeks GA:

•Start with 15 to 25 microW/cm²/nm to one (dorsal/ventral) body surface.

•Increase to 15 to 25 microW/cm²/nm to both ventral and dorsal body surfaces.

•Pre-exchange – 25 microW/cm²/nm to both ventral and dorsal body surfaces.

The use of phototherapy blanket allows exposure to both dorsal and ventral body surfaces. Infants who require this degree of intensive phototherapy often have a hemolytic disease and/or exposure to agents that may cause hemolysis. In these infants, other approaches, including exchange transfusion, may need to be considered. Because of concerns about the safety of phototherapy in extremely preterm infants, ongoing research is being conducted on alternative methods of providing phototherapy, including the use of lower irradiance levels and cycled phototherapy [47,48]. (See 'Cycled versus continuous phototherapy' below.)

Use of phototherapy in preterm neonates is discussed in greater detail below (see 'Phototherapy' below). Additional discussion of phototherapy more broadly is provided separately. (See "Unconjugated hyperbilirubinemia in term and late preterm newborns: Initial management", section on 'Initial intervention (phototherapy)'.)

Indications for exchange transfusion — Data supporting the indications for exchange transfusion are even more scant than that for phototherapy in preterm infants <35 weeks GA. As a result, there is a wide practice variation, and the decision to perform an exchange transfusion in a preterm infant is based upon the clinical judgment of the attending neonatologist. Assessment of the risk/benefit ratio of an exchange transfusion needs to include the safety, efficiency, and known complications of an exchange transfusion. We recommend performing an exchange transfusion in any infant with neurologic signs consistent with BIND, and suggest exchange transfusion for preterm infants with a TB target level that fails to respond to phototherapy.

An operational guideline based on the preterm infant's BW was developed using data from the 1974 NICHD phototherapy trial [41,42]. We use this stratified approach using GA to consider exchange transfusions in preterm infants without neurologic findings based on the NICHD data and published expert opinions, including the authors' [21]. For infants whose TB levels are approaching the threshold TB level for exchange transfusion, intensive phototherapy is administered. TB is measured every 4 to 8 hours to determine whether there is an adequate response, thus avoiding exchange transfusion.

Exchange transfusion is indicated based on GA and TB as follows (figure 2):

●GA <28 weeks – TB 11 to 14 mg/dL (188 to 239 micromol/L)

●GA 28 to 29 weeks – TB 12 to 14 mg/dL (205 to 239 micromol/L)

●GA 30 to 31 weeks – TB 13 to 16 mg/dL (222 to 274 micromol/L)

●GA 32 to 33 weeks – TB 15 to 18 mg/dL (257 to 308 micromol/L)

●GA >34 weeks – TB 17 to 19 mg/dL (291 to 325 micromol/L)

During the exchange transfusion, phototherapy is continued, although a single device may only be used. After completion of the exchange transfusion, phototherapy is restarted at the same dosing prior to procedure. TB is remeasured hours after the exchange transfusion is completed.

Other approaches — Examples of guidelines from other countries for the management of hyperbilirubinemia include those developed in the United Kingdom and Norway:

●The United Kingdom guideline to initiate phototherapy and exchange transfusion is based on consensus opinion and reported care practice in neonatal units (National Institute for Health and Care Excellence [NICE] guideline for neonatal jaundice). The threshold to initiate treatment for preterm infants greater than 72 hours of age is determined using the following formula that incorporates GA (weeks).

●The guideline to initiate phototherapy from Norway is based on the BW and postnatal age of the preterm infant [49]. In general, for infants >4 days of age, the following threshold levels of TB are used based on BW:

•BW >2500 g, but GA between 34 and 37 weeks – 17.5 mg/dL (300 micromol/L)

•BW between 1500 and 2500 g – 14.6 mg/dL (250 micromol/L)

•BW between 1000 and 1499 g – 11.7 mg/dL (200 micromol/L)

•BW between <1000 g – 8.8 mg/dL (150 micromol/L)

INTERVENTIONS

Phototherapy

Benefit and risk — Phototherapy is the most commonly used intervention to treat and prevent severe hyperbilirubinemia, including in preterm infants, as it has been shown to reduce total serum/plasma bilirubin (TB) [46]. Preterm infants born <35 weeks gestational age (GA) are at greater risk to develop bilirubin-induced neurologic dysfunction (BIND) than the more mature infants [8]. However, the prevalence of chronic bilirubin encephalopathy (CBE, previously referred to as kernicterus) has decreased to the point where CBE is rare. The use of phototherapy as outlined above has dramatically reduced the need for exchange transfusions, particularly in infants of birth weight <1500 g [50,51]. However, there are concerns of potential risks associated with phototherapy, including increased mortality in the most immature infants (birth weight [BW] <750 g). Nevertheless, based on the available evidence, we continue to use phototherapy using threshold bilirubin for different GAs to prevent and/or reduce the severity of BIND. (See 'Initiation of phototherapy' above.)

Phototherapy and mortality — Concerns have been raised that phototherapy may be harmful to extremely preterm (EPT) infants (GA <28 weeks). The proposed mechanisms are that phototherapy may cause direct photo-oxidative injury to cell membranes or it may indirectly cause injury by reducing bilirubin levels, which may play a protective antioxidant role. In either case, the most immature infants would be most susceptible as there is greater transmission of potentially toxic light through their thin, gelatinous skin, and they are at the greatest risk for oxidant injury because of reduced natural antioxidants.

However, data are conflicting regarding the association of phototherapy and mortality in EPT infants:

●A NICHD trial of 1974 extremely low birth weight (ELBW) (BW <1000 g) randomly assigned and stratified based on BW (low BW: 501 to 750 g and high BW: 751 to 1000 g) aggressive (initiated at a lower TB threshold) or conservative phototherapy [14]. Mean peak TB levels were lower using aggressive phototherapy (7 versus 9.8 mg/dL [120 versus 168 micromol/L]), and there were no differences in the primary composite outcome of death or neurodevelopmental impairment (NDI) at 18 to 22 months corrected age between aggressive and conservative therapy (52 versus 55 percent, relative risk [RR] 0.94, 95% CI 0.87-1.02), or death (24 versus 23 percent. RR 1.05, 95% CI 0.90-1.22).

However, a second published subanalysis reported that aggressive phototherapy compared with conservative phototherapy was associated with increased mortality for the most preterm infants (BW 501 to 750 g) who were mechanically ventilated and a reduction in NDI at 18 to 22 months corrected age [15].

●In contrast, a NICHD trial of 1802 patients (91.4 percent of the original cohort) of whom 1607 received phototherapy treatment (PTx) and 195 did not (NoPTx), reported phototherapy was not independently associated with the primary outcome at 18 to 22 months corrected age of death or NDI, death alone, or NDI [52]. In this cohort, the NoPTx group had a higher mean BW and GA, and rate of caesarean section, and was more likely to have mothers with a lower educational level. Subgroup analyses based on BWs reported a trend for higher mortality for infants with BWs between 501 and 750 for the NoPTX versus PTx group (47 versus 36 percent, odds ratio [OR] 0.58, 95% CI 0.32-1.04).

Technique and light sources — We suggest blue light-emitting diodes (LEDs) as the light source for phototherapy as they appear to achieve greater reduction in TB levels compared with other light sources in preterm infants [46]. In a study from the National Institute of Child Health and Human Development (NICHD) comparing different phototherapy devices in EPT infants, the absolute decrease in TB levels during the first 24 hours of therapy was greatest for devices with blue LEDs (2.2 mg/dL) followed by halogen spotlights (1.7 mg/dL), fluorescent light banks (1.3 mg/dL), and blankets (0.8 mg/dL) [46].

If fiberoptic blankets are used in ELBW infants, caution should be used to minimize risk of skin breakdown.

The procedure for administering phototherapy and selection of light sources are discussed in greater detail separately. (See "Unconjugated hyperbilirubinemia in term and late preterm newborns: Initial management", section on 'Administration'.).

Cycled versus continuous phototherapy — Because of the concern that phototherapy may contribute to an increase in mortality risk for EPT infants, it has been proposed that cycled rather than continuous phototherapy would lessen phototherapy exposure, thereby reducing side effects associated with prolonged light exposure, and still adequately control and reduce TB levels. A multicenter trial of EPT infants reported that cycled (≥15 min/hour) versus continuous phototherapy reduced phototherapy exposure (34 versus 72 hours, adjusted difference -40 hours, 95% CI -45 to -32 hours) with only a slight increase in mean peak TB (7.1 versus 6.4 mg/dL [121 versus 109 micromol/L], adjusted difference 0.7 mg/dL, 95% CI 0.4 to 1.1 mg/dL [12 micromol/L, 95% CI 7 to 19]) [53]. In a subset of patients, there was no difference in predischarge brainstem auditory-evoked response (BAER) wave V latency between the two treatment regimens. Based on these results, future larger trials are being planned to determine whether cycled phototherapy would improve survival without further NDI in EPT infants.

Prophylactic phototherapy — We suggest not using prophylactic phototherapy given the potential risks associated with phototherapy. Although there are some data to suggest prophylactic phototherapy would prevent hyperbilirubinemia [54], prophylactic use would needlessly expose some preterm infants to phototherapy who would otherwise never require phototherapy and potentially lead to increased costs and potential adverse outcomes for preterm infants.

Exchange transfusion — Data to support the indications for exchange transfusion are even more scant than those for phototherapy in infants <35 weeks GA. Information on the efficacy of exchange transfusion is extrapolated from observational data from term infants. Discussion on the benefit and risk of exchange transfusion and a description of the procedure are found separately. (See "Unconjugated hyperbilirubinemia in term and late preterm newborns: Escalation of care", section on 'Exchange transfusion'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Neonatal jaundice".)

SUMMARY AND RECOMMENDATIONS

●Importance – Infants exposed to elevated total serum/plasma bilirubin (TB) level (hyperbilirubinemia) are at-risk for bilirubin-induced neurologic dysfunction (BIND). Preterm infants, compared with term infants, appear to be more vulnerable to BIND at lower TB levels. In addition, hyperbilirubinemia in preterm infants is more prevalent and protracted than in term infants because of the immaturity of their red blood cells (RBCs), livers, and gastrointestinal tracts (figure 1). (See 'Bilirubin-induced neurologic dysfunction' above and 'Pathophysiology' above.)

●Screening – Screening for hyperbilirubinemia is based on the rationale that early identification and treatment of hyperbilirubinemia will reduce the risk of BIND. In our center, we screen all preterm neonates admitted to the neonatal intensive care unit with a TB level in the first 24 hours after birth, with subsequent testing performed every 24 hours. The frequency of testing increases as TB level approach thresholds for interventions. (See 'Screening for hyperbilirubinemia' above.)

●Clinical manifestations – Clinical findings of hyperbilirubinemia are due to bilirubin deposition in the skin (jaundice) and the brain. (See 'Clinical presentations' above.)

•Jaundice is the yellowish discoloration of the skin, subcutaneous tissue, and/or conjunctiva (as visualized on the sclerae) caused by bilirubin deposition. (See 'Jaundice' above.)

•In preterm infants <35 weeks gestational age (GA), the acute neurologic signs (eg, hypertonia, irritability, posturing, arching, and seizures) commonly seen in more mature neonates are not typically observed. Preterm infants may only manifest minimal findings of recurrent apnea, periodic breathing, or episodes of oxygen desaturation, which may be masked in mechanically ventilated infants. Automated auditory brainstem-evoked response (AABR) may detect acute reversible bilirubin-induced auditory dysfunction. (See 'Acute bilirubin encephalopathy (ABE)' above.)

•Chronic bilirubin encephalopathy (CBE, previously referred to as kernicterus), is the chronic form of BIND associated with permanent neurologic sequelae. The clinical manifestations of CBE in preterm infants are similar to those observed in term infants and include choreoathetoid cerebral palsy, sensorineural hearing loss, upward gaze abnormalities, and enamel dysplasia of deciduous teeth. (See 'Chronic bilirubin encephalopathy (CBE) or kernicterus' above.)

●Management approach – A threshold TB level that is predictive of neurologic sequelae in preterm infants has yet to be identified. As a result, consensus is lacking on the optimal management hyperbilirubinemia in infants <35 weeks GA. Our suggested approach is as follows (see 'Management approach' above):

•Phototherapy – For preterm infants <35 weeks GA and <7 days old, we suggest starting phototherapy based upon the following GA and TB thresholds (figure 2) (Grade 2C)(see 'Initiation of phototherapy' above and 'Benefit and risk' above):

-GA <28 weeks – TB >5 mg/dL (86 micromol/L)

-GA 28 to 29 weeks – TB 6 to 8 mg/dL (103 to 137 micromol/L)

-GA 30 to 31 weeks – TB 8 to 10 mg/dL (137 to 171 micromol/L)

-GA 32 to 33 weeks – TB 10 to 12 mg/dL (171 to 205 micromol/L)

-GA 34 to <35 weeks – TB 12 to 14 mg/dL (205 to 239 micromol/L)

Other centers base decisions on TB levels stratified by birthweight. We suggest blue light-emitting diodes (LEDs) rather than other light sources for phototherapy (Grade 2C) since they appear to achieve greater reduction in TB levels. Phototherapy is initially applied to the ventral body area with uniform exposure (picture 1) at an irradiance of 15 to 30 microW/cm² per nm depending on GA. (See 'Technique and light sources' above.)

•Exchange transfusion – For preterm infants <35 weeks GA with hyperbilirubinemia that does not respond adequately to intensive phototherapy and that exceeds the TB threshold level based on GA (figure 2), we suggest performing exchange transfusion (Grade 2C). (See 'Indications for exchange transfusion' above.)