Postnatal weight loss in term infants: what is "normal" and do growth charts allow for it?

BACKGROUND: Although it is a well known phenomenon, limited normative data on neonatal weight loss and subsequent gain are available, making it hard to assess individual children with prolonged weight loss. OBJECTIVE: To establish, using data from a large prospective population based cohort study, norms and limits for postnatal weight loss and its impact on current growth reference charts. METHOD: A cohort of 961 term infants were recruited at birth and followed using parental questionnaires and community nursing returns. Routine weights were collected for half the cohort at 5 days and for all at 12 days and 6 weeks. RESULTS: Less weight loss was seen than the 3-6% suggested by previous studies, but one in five infants had not regained their birth weight by 12 days. Those lightest at birth showed least weight loss. Twenty six (3%) children had more than 10% weight loss, but none showed evidence of major organic disease. Actual weights in the first fortnight are half to one centile space lower than growth charts suggest, while birthweight centiles for children born at 37 weeks were two centile spaces lower. CONCLUSIONS: Neonatal weight loss is brief, with few children remaining more than 10% below birth weight after 5 days. Growth charts are misleading in the first 2 weeks, because they make no allowance for neonatal weight loss.     

Weight reference charts for British long-term breastfed infants

Aim: To produce a modification of the British 1990 weight reference reflecting the growth of long-term breastfed infants. Methods: 120 infants from the Cambridge Infant Growth Study fed breast milk (with no formula) for at least 24 wk, with solids introduced at a mean age of 15 wk, were weighed every 4 wk from birth to 52 wk. Weights were converted to standard deviation scores (SDS) for gender and age post-term based on the revised British 1990 reference, averaged and smoothed across age, and then converted back to weights to provide modified median weight curves by gender. Other centile curves were constructed assuming the same variability and skewness by age as for the British reference. Results: Long-term breastfed infants were slightly heavier than the reference at birth and crossed centiles upwards to reach #0.3 SDS at 2 mo, but subsequently crossed centiles downwards to 30.2 SDS by 12 mo. Conclusion: The British 1990 reference, although better than previous growth standards, reflects the growth of long-term breastfed infants only imperfectly, with mean weight falling by 0.5 SDS from 2 to 12 mo. The provision of breastfeeding-specific centiles, either as a transparent overlay or as an alternative chart, should be useful for professionals and parents to monitor the weight of long-term breastfed infants.     

Growth and very low birth weight.

Information on the likelihood of catch up growth in poorly grown very low birthweight children is sparse. The centiles for weight, height, and head circumference were recorded at both 2 and 5 years of age for 135 very low birthweight children and 42 normal birthweight children. At both ages significantly more children of very low birth weight were under the 10th centile for weight and height. Children of birth weight under 1000 g were more often under the 10th centile for weight at 5 years compared with those of birth weight 1000-1500 g. Mean incremental weight gain between 2 and 5 years was significantly less for very low birthweight children. Mean increment in weight from 2 to 5 years was less for very low birthweight children who had been under the 10th centile for weight at 2 years; children who had been under the 10th centile for height also had lower mean height increments. The growth centiles achieved by 2 years of age were useful predictors of poor growth at 5 years, with perinatal data of marginal importance. Only six of 43 (14%) children with a weight at 5 years of age under the 10th centile were small for gestational age at birth. Very low birthweight children who had a weight or height under the 10th centile at 2 years of age usually remained in this category at 5 years with no evidence of catch up growth.     

Influence of folic acid on birthweight and growth of the erythroblastotic infant. III. Effect of folic acid supplementation.

Seventeen infants with severe (14) or moderately severe (3) erythroblastosis were given daily oral supplements of 2-5 or 5 mg folic acid from day 16 (average) to 3-2 months (average). Their rate of weight gain, expressed as weight centiles, was followed for 1 year and was compared with that of a very similar group of 34 erythroblastotic infants without folic acid supplements. By the end of the 4th month, just after stopping additional folate intake, the median centiles for weight had risen from the 40th to the 80th centile, while in the untreated control group they rose during this period from the 35th to the 50th centile. During the second half of the year both groups declined in weight centiles, the ''treated'' group ending up at the 50th centile for weight, while the control group fell to the 25th.     

Ten-year trends in obesity among Israeli schoolchildren: 1990-2000

AIM: To analyze trends in childhood body mass index (BMI) in Israel between 1990 and 2000, and to determine the proportion of obese children using US and population-specific reference values. METHODS: Cross-sectional data from 13 284 second- and fifth-grade schoolchildren were collected, including age, sex, height, weight, country of birth, and time since immigration. Age- and sex-specific BMI means and centiles were calculated, and the prevalence of obesity was determined using Israeli and US reference values. RESULTS: BMI values at the 95th centile increased monotonously over time in all age and sex categories. Between 1990 and 2000, 95th centile values increased by 12.7% and 11.8% among second-grade boys and girls, respectively. Among fifth-grade children, 95th centile values increased by 10.2% and 8.4%, respectively. Among second graders in 2000, 11.4% of both boys and girls exceeded the BMI value recorded at the 95th centile in 1990. Among fifth-graders in 2000, 10.7% of boys and 11.1% of girls exceeded the 1990 BMI reference value (p for all comparisons < 0.001). The proportion of obese children increased over time using both Israeli and US reference values. CONCLUSION: This substantial increase in childhood obesity poses a serious health threat, and requires implementation of suitable public health interventions.     

Procalcitonin in preterm infants during the first few days of life: introducing an age related nomogram

OBJECTIVE: To determine normal concentrations of procalcitonin in preterm infants shortly after birth and to assess its accuracy in detecting bacterial infection. METHODS: Blood samples of 100 preterm infants were prospectively drawn during the first 4 days of life for determination of procalcitonin concentration. Infants were classified into four groups according to their sepsis status. RESULTS: Mean (SD) gestational age and birth weight were 32 (2.9) weeks and 1682 (500) g respectively. A total of 283 procalcitonin concentrations from healthy infants were plotted to construct nomograms of physiologically raised procalcitonin concentration after birth, stratified by two groups to 24-30 and 31-36 weeks gestation. The peak 95th centile procalcitonin concentration was plotted at 28 hours of age; values return to normal after 4 days of life. Only 12 infants were infected, and 13 of their 16 procalcitonin concentrations after birth were higher than the 95th centile, whereas samples taken at birth were lower. In a multivariable analysis, gestational age, premature rupture of membrane, and sepsis status influenced procalcitonin concentration independently, but maternal infection status did not. CONCLUSIONS: The suggested neonatal nomograms of preterm infants are different from those of term infants. Procalcitonin concentrations exceeding the 95th centile may be helpful in detecting congenital infection, but not at birth.     

Customised birthweight standards accurately predict perinatal morbidity

OBJECTIVE: Fetal growth restriction is associated with adverse perinatal outcome but is often not recognised antenatally, and low birthweight centiles based on population norms are used as a proxy instead. This study compared the association between neonatal morbidity and fetal growth status at birth as determined by customised birthweight centiles and currently used centiles based on population standards. DESIGN: Retrospective cohort study. SETTING: Referral hospital, Barcelona, Spain. PATIENTS: A cohort of 13 661 non-malformed singleton deliveries. INTERVENTIONS: Both population-based and customised standards for birth weight were applied to the study cohort. Customised weight centiles were calculated by adjusting for maternal height, booking weight, parity, ethnic origin, gestational age at delivery and fetal sex. MAIN OUTCOME MEASURES: Newborn morbidity and perinatal death. RESULTS: The association between smallness for gestational age (SGA) and perinatal morbidity was stronger when birthweight limits were customised, and resulted in an additional 4.1% (n=565) neonates being classified as SGA. Compared with non-SGA neonates, this newly identified group had an increased risk of perinatal mortality (OR 3.2; 95% CI 1.6 to 6.2), neurological morbidity (OR 3.2; 95% CI 1.7 to 6.1) and non-neurological morbidity (OR 8; 95% CI 4.8 to 13.6). CONCLUSION: Customised standards improve the prediction of adverse neonatal outcome. The association between SGA and adverse outcome is independent of the gestational age at delivery.     

Recognising failure to thrive in early childhood.

The maximum weight centile achieved by a child between 4 and 8 weeks of age was found to be a better predictor of the centile at 12 months than the birth weight centile. Children whose weight deviated two or more major centiles below this maximum weight centile for a month or more showed significant anthropometric differences during the second year of life from those who showed no such deviation. It is suggested that this leads to a logical and practical definition of failure to thrive.     

Prenatal predictors of mortality in very preterm infants cared for in the Australian and New Zealand Neonatal Network

AIM: To identify antenatal and perinatal risk factors for in-hospital mortality of babies born within the Australian and New Zealand Neonatal Network (ANZNN). METHODS: Data were collected prospectively as part of the ongoing audit of high-risk infants (birth weight <1500 g or gestation <32 weeks) admitted to all level III neonatal units in Australia and New Zealand. Antenatal and intrapartum factors to 1 min of age were examined in 11 498 infants with gestational age >24 weeks. Risk and protective factors for mortality were derived from logistic regression models fitted to 1998-9 data and validated on 2000-1 data. RESULTS: For the whole cohort of infants born between 1998 and 2001, prematurity was the dominant risk factor, infants born at 25 weeks having 32 times greater odds of death than infants born at 31 weeks. Low birth weight for gestational age also had a dose-response effect: the more growth restricted the infant the greater the risk of mortality; infants below the 3rd centile had eight times greater odds of death than those between the 25th and 75th centiles. Male sex was also a significant risk factor (odds ratio (OR) 1.55, 95% confidence interval (CI) 1.31 to 1.82). Maternal hypertension in pregnancy was protective (OR 0.46, 95% CI 0.36 to 0.50). The predictive model for mortality had an area under the receiver operating characteristic curve of 0.82. CONCLUSIONS: Risk of mortality can be predicted with good accuracy with factors up to the 1 min Apgar score. By using gestation rather than birth weight as the main indicator of maturity, these data confirm that weight for gestational age is an independent risk factor for mortality.     

High faecal calprotectin concentrations in newborn infants

BACKGROUND: Calprotectin, a major component of soluble cytosolic proteins in human neutrophil granulocytes, is excreted in excess in stools during inflammatory bowel disease in adults and children. Faecal calprotectin concentrations are also higher during the first year of life than in adults. OBJECTIVES: To measure faecal calprotectin concentrations in the neonatal period and define reference values according to the mode of feeding: standard infant formula, prebiotic infant formula (Calisma, Blédina SA, France), or breast feeding.Patients and methods: A prospective study was carried out over three months in 69 full term, healthy newborns with a median gestational age of 39.8 weeks (range 37-41.5) and a birth weight of 3300 g (range 2600-4460). Three groups were formed depending on the mode of feeding: group 1 (n = 18) received a standard infant formula, group 2 (n = 19) the prebiotic infant formula, and group 3 (n = 32) was breast fed. One stool sample was taken from each newborn on day 4 (3-7), and faecal calprotectin analysed using a commercial enzyme linked immunoassay (Calprest, Eurospital, Italy). RESULTS: Faecal calprotectin concentrations (median 167 micro g/g) were higher than reference values in healthy adults. The concentration was below the upper reference limit for adults (50 micro g/g) for three infants only, one fed the standard formula and two fed the prebiotic formula. Concentrations did not differ significantly according to method of feeding. CONCLUSIONS: Compared with healthy adults, newborns have high calprotectin concentrations in the first days of life. There is no obvious influence of the mode of feeding.     

Blood visfatin concentrations in normal full-term pregnancies

AIM: To prospectively investigate blood visfatin concentrations during the perinatal period in normal pregnancies. METHODS: Visfatin concentrations were determined in maternal, umbilical cord (representing the foetal state) and neonatal blood on day 1 (N1) and 4 (N4). RESULTS: Maternal and foetal visfatin concentrations were similar (18.83+/-4.27 and 19.35+/-4.90 ng/mL, respectively). There were significant correlations between maternal and foetal (r=0.742, p<0.001), as well as between N1 and N4 (r=0.487, p=0.029) visfatin concentrations. Foetal concentrations were significantly elevated compared to N1 (p=0.032). There was no difference between N1 and N4 concentrations. However, there was a correlation between birth weight and neonatal visfatin concentrations: there was a mean increase in N1 and N4 visfatin concentrations by 0.221 ng/mL and 0.292 ng/mL, respectively, for every unit increase in customized centile (adjusted birth weight) (p=0.021 and p=0.005, respectively). No association was found between serum visfatin concentrations and gender, parity or mode of delivery. CONCLUSIONS: Expression in foetal membranes and placental transfer could be responsible for higher blood visfatin concentrations during intrauterine life. Customized centiles seem to be independent predictor variables for postnatal visfatin concentrations. This finding could be attributed to the production of visfatin in adipose tissue, a main contributor to birth weight and consequently to customized centiles.     

Randomized, double-blind comparison of growth in infants receiving goat milk formula versus cow milk infant formula

OBJECTIVE: To compare growth of infants fed goat milk infant formula (GMF) or cow milk infant formula (CMF) and to compare tolerability and safety of the two formulas. METHODS: The study was conducted in Auckland, New Zealand. This was a double-blind randomized controlled trial. Newborn term infants were randomized within 72 h of birth to GMF or CMF. Milk formula powder in single serve sachets were reconstituted and fed to infants from trial commencement until age 168 days. No other formula given from randomization until age 168 days. Infant weight, length and head circumference were measured at birth and age 14, 28, 56, 84, 112, 140 and 168 days. Bowel motion frequency and consistency, sleeping and crying patterns and adverse events were also measured. RESULTS: Seventy-two infants were randomized, 36 each to GMF or CMF, with 62 infants completing the intervention. At enrollment the average weight of infants in the GMF group (mean +/- SD) was 3.33 +/- 0.43 kg and in the CMF group 3.43 +/- 0.47 kg; and at study completion 8.07 +/- 0.90 kg (GMF) and 7.87 +/- 0.99 kg (CMF). The difference in average weight gain over the study period for the GMF group versus the CMF group was not significant (+309 g; 95% CI = -49 to +668, P = 0.09). Median daily bowel motion frequency was greater in the GMF group than the CMF group (2.4 vs 1.7, P = 0.01). There were no group differences in bowel motion consistency, duration of crying, ease of settling, or frequency of adverse events. CONCLUSION: Growth of infants fed GMF is not different to that of infants-fed CMF.     

Motor skills in adolescents with low birth weight

BACKGROUND: Minor motor problems have been reported in low birthweight children, but few studies have assessed motor skills in adolescents. OBJECTIVE: To examine the prevalence of motor problems in adolescents with low birth weight. METHOD: Fifty four very low birthweight (VLBW: birth weight < or = 1500 g), 59 term small for gestational age (SGA: birth weight < 10th centile), and 83 control (birth weight > or = 10th centile at term) children were assessed with the Movement assessment battery for children (Movement ABC) at the age of 14 in a population based study. RESULTS: One in four VLBW children (odds ratio (OR) 9.3, 95% confidence interval (CI) 2.5 to 34.5) and one in six SGA children (OR 4.7, 95%CI 1.2 to 18.4) had motor problems compared with controls (3.7%). There were no sex differences in motor problems in the VLBW group, and the increased risk was consistent across the continuum of the Movement ABC. For SGA children, the increased risk of motor problems was particularly in manual dexterity in boys. CONCLUSION: VLBW and SGA adolescents have increased risk of motor problems compared with control children.     

Implications of adopting the WHO 2006 Child Growth Standard in the UK: two prospective cohort studies.

BACKGROUND: The WHO 2006 Child Growth Standard is based on data from international optimally nourished breastfed infants from birth to age 5 years. OBJECTIVE: To assess the potential effect of its use on weight and growth monitoring of UK children. PARTICIPANTS: Full-term members of two population-based UK birth cohorts: the Children in Focus sub-cohort of the Avon Longitudinal Study of Parents and Children (ALSPAC) (n = 1335) and the Gateshead Millennium Baby Study (GMS; n = 923). DESIGN: Growth data from birth to 5 years were converted into z-scores relative to the WHO 2006 standard. RESULTS: Compared with the WHO standard, both UK cohorts had higher birth weights (mean z-scores: GMS, 0.17; ALSPAC, 0.34) and ALSPAC had higher birth lengths. After birth, length showed a good fit at all ages. By 2-4 months, both cohorts were similar in weight to the WHO median (mean WHO weight z-score at 4 months: GMS, 0.01; ALSPAC, -0.07), but thereafter the UK cohorts were heavier (mean WHO weight z-score at 12 months: GMS, 0.57; ALSPAC, 0.65). At age 12 months, the risk of being classified as underweight (weight <2nd centile) was considerably lower according to the WHO standard than by the UK 1990 Growth Reference (RR = 0.15, 95% CI = 0.07 to 0.32), and the risk of being classified as obese at 4-5 years (body mass index >98th centile) was slightly increased (RR = 1.35, 95% CI = 1.02 to 1.78). CONCLUSIONS: Adoption of the WHO 2006 Growth Charts would set a markedly lower standard of weight gain beyond the age of 4 months for UK infants and could support efforts to avoid future childhood obesity. However, the WHO standard is not representative of size at birth in the UK.     

The Wessex Growth Study: First Report

Voss, L., Walker, J., Lunt, H., Wilkin, T. and Betts, P. (Department of Paediatrics and the Endocrine Section, Department of Medicine II, General Hospital, Southampton, UK). The Wessex growth study: first report. Acta Paediatr Scand [Suppl] 349: 65, 1989.
The Wessex Growth Study is a community-based longitudinal survey of short children recruited from two Health Districts in Wessex during 1985–86 (cohort I) and 1986–87 (cohort II). Screening of new school entrants during 1985–86 identified only 1.3% who were at or below the 3rd centile for height as defined by Tanner and Whitehouse, suggesting a strong secular trend and an urgent need for updated height charts. After exclusion of the small number of children with underlying organic pathology and those from ethnic minorities, there were 84 children in cohort I on whom this report is based. These apparently normal, short children were sex- and age-matched with normal controls (10th-90th centile) from the same school class. Forty-two per cent of cohort I had a delayed bone age, and 34% lay above the 3rd centile after correction for parental height. Forty-four per cent were of low birth weight. The correlation between two successive height velocities measured at 6 and 12 months was only -0.14 for cohort I and -0.15 for their controls. Twelve-month mean height velocity SD scores (SDS) were -0.45 and +0.25, respectively, corresponding to the 33rd and 60th centiles, and rather higher than the 25th and 50th centiles expected in view of the heights of these children. Thirty-eight per cent of cohort 1 had a 12-month height velocity below the 25th centile, and in 16% height velocity was below the 10th centile. As a group, the children in cohort I grew more slowly than their controls, hut the height velocity in 88% of cases lay within the control range. For any individual, therefore, 12-month height velocity was not a useful discriminant of growth failure. Repeated measurements are indicated.     

Supplementary feeding in the maternity ward shortens the duration of breast feeding

In a prospective study, feeding routines of a maternity unit and the subsequent feeding patterns of 521 newborns were analysed. During the stay in the maternity unit, 69% of newborns were exclusively breast fed and 1% received only donor's milk from the milk bank and/or formula. Nine percent received their mothers' milk by bottle at least once and 21% received one or more supplementary feedings with donor's milk from the milk bank. One-quarter of the children received supplementary feeds on the third day of life, the indications for this being birth weight less than 3.0 kg, maternal diabetes or gestational diabetes, “insufficient amounts” of milk or fussiness. At three months, 65% were being exclusively breast fed and 15% partially breast fed. In a multiple logistic regression analysis, the potential determinants (neonatal feeding, maternal characteristics, characteristics of the delivery and the child) for the duration of breast feeding were included. The adjusted relative risk (estimated as odds ratios, OR) of not being breast fed at three months was associated with maternal age (< 25 years, OR 4.2), maternal smoking (OR 4.0), neonatal feeding (supplements given, OR 3.9) and initial weight loss (10% or more, OR 2.8). Thus the administration of supplementary donor's milk or formula during the early neonatal period was associated with an increased risk of a short duration for breast feeding, even after adjustment for a number of potential confounders.     

Identifying newborns at risk of significant hyperbilirubinaemia: a comparison of two recommended approaches.

AIMS: To compare the predictive performance of clinical risk factor assessment and pre-discharge bilirubin measurement as screening tools for identifying infants at risk of developing significant neonatal hyperbilirubinaemia (post-discharge total serum bilirubin (TSB) >95th centile). METHODS: Retrospective cohort study of term and near term infants born in an urban community teaching hospital in Pennsylvania (1993-97). A clinical risk factor scoring system was developed and its predictive performance compared to a pre-discharge TSB expressed as a risk zone on a bilirubin nomogram. Main outcome measures were prediction model discrimination, range of predicted probabilities, and sensitivity, specificity, positive and negative predictive values, and likelihood ratios for various positivity criteria. RESULTS: The clinical risk factor scoring system developed included birth weight, gestational age <38 weeks, oxytocin use during delivery, vacuum extraction, breast feeding, and combination breast and bottle feeding. The pre-discharge bilirubin risk zone had better discrimination (c = 0.83; 95% CI 0.80 to 0.86) than the clinical risk factor score (c = 0.71; 95% CI 0.66 to 0.76) and predicted risk of significant hyperbilirubinaemia as high as 59% compared with a maximum of 44% for the clinical risk factor score. Neither the risk score nor the pre-discharge TSB risk zone predicted the outcome with > or =0.98 sensitivity without significantly compromising specificity (0.13 and 0.21, respectively). Multi-level clinical risk factor scores and TSB risk zones produced likelihood ratios of 0.15-3.25 and 0.05-9.43, respectively. CONCLUSIONS: The pre-discharge bilirubin expressed as a risk zone on an hour specific bilirubin nomogram is more accurate and generates wider risk stratification than a clinical risk factor score.     

Perinatal predictors of growth patterns to 18 months in children born small for gestational age

AIM: To determine factors before or at birth that are predictive of growth patterns to 18 months in children born small for gestational age (SGA). METHODS: Prospective cohort study of 186 SGA babies. Catch-up growth patterns were defined as early (>10th centile at 6 and 18 months), late (<10th centile at 6 months but >10th centile at 18 months), transient (>10th centile at 6 months but <10th centile at 18 months) or none (<10th centile at 6 and 18 months). RESULTS: Most children (75%) showed catch-up growth by 6 months. Of antenatal variables studied, only early gestation at diagnosis of SGA predicted late or failed catch-up. Late or failed catch-up was also associated with short gestation, small absolute and relative size at birth, increased placental weight/birthweight ratio (Pl/BW) and prolonged neonatal hospital stay. On logistic regression, both late and failed catch-up were associated with short birth length. Late catch-up growth was also associated with prolonged hospital stay and male sex. Failed catch-up was associated with increased Pl/BW. No antenatal or perinatal variables distinguished early from transient catch-up groups. CONCLUSION: SGA babies with late onset and less severe growth restriction have a good chance of catch-up growth by 6 months of age. Catch-up growth is likely to be delayed in SGA babies who are short at birth, are boys, and have prolonged hospital stays. However, poor growth over the first 6 months does not predict later growth patterns. Failure of catch-up growth in babies with increased Pl/BW may reflect an intrinsic growth defect. Transient catch-up growth may reflect environmental factors operating after birth.     

Prevalence and independent risk factors for hearing loss in NICU infants

AIM: To determine the prevalence and independent relationship between hearing loss and risk factors in a representative neonatal intensive care unit (NICU) population. METHODS: Automated auditory brainstem response (AABR) hearing screening has been introduced since 1998 in the Dutch NICUs. After a second AABR failure, diagnostic ABR was used to establish diagnosis of hearing loss. Newborns who died before the age of 3 months were excluded. In the present study only the NICU infants who were born with a gestational age <30 weeks and/or a birth weight <1000 g between October 1, 1998 and January 1, 2002 were included. Risk factors included in the study were familial hearing loss, in utero infections, craniofacial anomalies, birth weight <1500 g, hyperbilirubinemia, ototoxic medications, cerebral complications, severe birth asphyxia, assisted ventilation > or =5 days and syndromes. RESULTS: A nationwide cohort of 2186 newborns were included. Mean gestational age was 28.5 weeks (SD 1.6) and mean birth weight was 1039 g (SD 256). Prevalence of uni- or bilateral hearing loss was 3.2% (71/2186; 95% CI 2.6-4.1). Multivariate analysis revealed that the only independent risk factors for hearing loss were severe birth asphyxia (OR 1.7; 95% CI 1.0-2.7) and assisted ventilation > or =5 days (OR 3.6; 95% CI 2.1-6.0). CONCLUSION: The prevalence of hearing loss in a representative NICU population was 3.2%. Independent risk factors for hearing loss were severe birth asphyxia and assisted ventilation > or =5 days.     

Identifying newborns at risk of significant hyperbilirubinaemia: a comparison of two recommended approaches

Aims: To compare the predictive performance of clinical risk factor assessment and pre-discharge bilirubin measurement as screening tools for identifying infants at risk of developing significant neonatal hyperbilirubinaemia (post-discharge total serum bilirubin (TSB) >95th centile). Methods: Retrospective cohort study of term and near term infants born in an urban community teaching hospital in Pennsylvania (1993–97). A clinical risk factor scoring system was developed and its predictive performance compared to a pre-discharge TSB expressed as a risk zone on a bilirubin nomogram. Main outcome measures were prediction model discrimination, range of predicted probabilities, and sensitivity, specificity, positive and negative predictive values, and likelihood ratios for various positivity criteria. Results: The clinical risk factor scoring system developed included birth weight, gestational age <38 weeks, oxytocin use during delivery, vacuum extraction, breast feeding, and combination breast and bottle feeding. The pre-discharge bilirubin risk zone had better discrimination (c = 0.83; 95% CI 0.80 to 0.86) than the clinical risk factor score (c = 0.71; 95% CI 0.66 to 0.76) and predicted risk of significant hyperbilirubinaemia as high as 59% compared with a maximum of 44% for the clinical risk factor score. Neither the risk score nor the pre-discharge TSB risk zone predicted the outcome with ?0.98 sensitivity without significantly compromising specificity (0.13 and 0.21, respectively). Multi-level clinical risk factor scores and TSB risk zones produced likelihood ratios of 0.15–3.25 and 0.05–9.43, respectively. Conclusions: The pre-discharge bilirubin expressed as a risk zone on an hour specific bilirubin nomogram is more accurate and generates wider risk stratification than a clinical risk factor score.