This version was machine-translated from German and has not been medically reviewed. The authoritative version is the German version.

AAP Bilirubin 2004 (Legacy)

Older nomograms (low, intermediate, high risk)

Legacy calculator. The AAP guideline 2004 was superseded in 2022; it published its thresholds only as curves (“the levels shown are approximations”). The figures used here are a documented digitization (The original guideline shows the thresholds ONLY as curve graphs (Fig. 3/4) or percentile zones (Fig. 2); no official coordinate/number table exists for them (Appendix 1, 'Table 4' of the guideline is a different table – prevalence of TSB values per risk zone in the validation study, NOT curve support points). The numerical values stored here come from an openly licensed digitization: Rosenberg D. 'Biligram' (JavaScript tool for visualizing the Bhutani nomogram and the AAP phototherapy/exchange thresholds), GitHub github.com/dmr001/biligram, GNU GPL v3, file js/biligram.js, objects 'treatmentZone' (phototherapy/exchange, lines ~108-120) and 'riskZone' (Bhutani percentile thresholds, lines ~32-43); permalink: https://github.com/dmr001/biligram/blob/d9da2f6a7e784a69e599b448b60885068c81068a/js/biligram.js#L32-L120 . The project's help.html (https://github.com/dmr001/biligram/blob/master/help.html) explicitly names the original guideline (Pediatrics 2004;114:297) and the update of 2009 (Pediatrics 2009;124:1193) as the source of the curves; a detailed digitization methodology (e.g. software used) is not described there. We extracted the values programmatically from the source code (no manual typing). Additional verification: we viewed the original figures 2, 3 and 4 directly in a full-text PDF mirror of the original publication (identical page numbers/DOI 10.1542/peds.114.1.297 as Pediatrics 2004;114:297-316) – https://emergency.med.ufl.edu/files/2013/02/management-of-hyperbilirubinemia.pdf, p. 301/304/305 – and compared the Biligram support points by eye at the main grid lines (birth, 24/48/72/96/120/144 h or 12-144 h for Fig. 2) with the course of the curve: agreement throughout within about +/-0.5 mg/dL (visual estimate, no pixel measurement), which is within the usual reading accuracy of a hand-drawn curve. A second, independent digitization attempt (polynomial fit in the project smart-on-fhir-reference-implementation/bilirubin-risk-chart, js/services.js) was consulted for comparison but discarded (see notes) – it is NOT used here as a data source.) – Reliability: medium - There is NO official number table from the AAP for Fig. 2-4 (only curve graphs, described by the guideline itself as 'approximations'; the only table in Appendix 1 - Table 4 - concerns something else, see 'tabulation'). The stored values therefore come from an openly licensed (GPL-3), physician-authored third-party digitization (Biligram, Daniel Rosenberg MD), which explicitly names the AAP-2004 guideline as its source but does not document its digitization methodology. We spot-checked the values by visual eyeball comparison with the original figures (from a verified full-text PDF mirror) at the main grid lines (deviation approx. <= 0.5 mg/dL), which strengthens the fidelity of the digitization but does not replace an official table – hence 'medium' rather than 'high' (unlike aap2022.json in this folder, which is based on official AAP supplemental tables and deserves 'high'). Sufficiently reliable for clinical use in a calculator portal, but to be communicated with this limitation..

Child

Gestational age at birth

Birth (date, time)

required if measurements are entered with a time point

Unit of the values

mg/dl µmol/l

Risk factors (AAP 2004)

Isoimmune hemolysis, G6PD deficiency, asphyxia, marked lethargy, temperature instability, sepsis, acidosis or albumin < 3.0 g/dl

Bilirubin values

Total bilirubin	Time of measurement	… or hours of life

+ Add row

Reset

About this calculator

Phototherapy and exchange transfusion thresholds of the AAP guideline 2004 (Subcommittee on Hyperbilirubinemia) for the three risk groups – for comparison with the thresholds of 2022.

Bhutani zones (Bhutani VK et al. Pediatrics 1999;103:6–14) according to the hour-specific percentile of bilirubin before discharge, provided the curves are available in tabulated form.

Source

American Academy of Pediatrics Subcommittee on Hyperbilirubinemia. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics. 2004;114(1):297–316.

Notes on data origin

Risk groups defined exactly as in Fig. 3/4 of the guideline: 'low' = Infants at lower risk (>=38 wk and well); 'medium' = Infants at medium risk (>=38 wk + risk factors OR 35-37 6/7 wk. and well); 'high' = Infants at higher risk (35-37 6/7 wk. + risk factors). Risk factors per Fig.3/4 footnote: isoimmune haemolytic disease, G6PD deficiency, asphyxia, marked lethargy, temperature instability, sepsis, acidosis, albumin < 3.0 g/dL (if measured).
The original footnote to Fig. 3 reads verbatim: 'Note: These guidelines are based on limited evidence and the levels shown are approximations.' An explicit numeric footnote of the form '24 h: 12/10/8 mg/dL' or similar, as mentioned in the assignment as a verification note, could NOT be found in the original source - the footnotes contain only qualitative remarks (uncertainty in the first 24 h, definition of risk factors, option for individual adjustment at 35-37 6/7 weeks, B/A ratio table). The values actually digitized at 24 h (11.5/9.8/7.8) and 72 h (17.6/15.4/13.4) are, however, close to the approximate values named in the assignment (12/10/8 and 18/15/13 respectively), deviation <= 0.5 mg/dL - presumably roughly rounded rule-of-thumb values from clinical parlance, not from the guideline text itself.
The hour 0 value ('Birth') and the first 24 h are shown in Fig. 3 as dotted/dashed lines; Fig. 4 additionally notes verbatim: 'The dashed lines for the first 24 hours indicate uncertainty due to a wide range of clinical circumstances and a range of responses to phototherapy.' Values before 24 h are therefore subject to greater uncertainty than later values – this is a property of the original curve, not a digitization error.
Hour 168 (7 days) is present as a grid line in Fig. 3/4 of the original (x-axis extends to '7 Days'); the curves there are visually already on the plateau. The source digitization (Biligram) ends at 144 h; the value at 168 h was therefore added as a continuation of the plateau value at 144 h (no new curve information, but supported by visual inspection of the original graph).
Fig. 2 (Bhutani risk zones) of the AAP 2004 guideline is itself an adapted reproduction of Bhutani VK, Johnson LH, Sivieri EM. Predictive ability of a predischarge hour-specific serum bilirubin for subsequent significant hyperbilirubinemia in healthy term and near-term newborns. Pediatrics 1999;103:6-14 ('Used with permission from Bhutani et al'). The three digitized limit curves (p95/p75/p40) correspond to the limits 'High Risk Zone / High Intermediate Risk Zone / Low Intermediate Risk Zone / Low Risk Zone' from Fig. 2. Fig. 2 itself explicitly labels only the uppermost curve as '95th %ile'; the assignment of the middle/lower limit line to P75 and P40 respectively does not follow a label in the AAP figure but the zone definition from Bhutani et al. 1999 (<P40, P40-P75, P75-P95, >P95), to which Fig. 2 refers by footnote.
A second independent attempt to digitize the Bhutani curves (polynomial of degree 6 in github.com/smart-on-fhir-reference-implementation/bilirubin-risk-chart, js/services.js) was examined but rejected: it deviates markedly at 24 h (p95 approx. 9.8 instead of 7.8 mg/dL) and diverges unrealistically beyond approx. 130 h (e.g. 113 mg/dL at 168 h) - a typical artifact of a globally fitted high-degree polynomial. This source was NOT used to determine values, only to assess uncertainty.
The 'medium' group combines two clinically different populations on one curve (>=38 weeks + risk factors OR 35-37 6/7 weeks and healthy). According to the Fig.3 footnote, the guideline itself allows fine adjustment here: 'For well infants 35-37 6/7 wk we can adjust TSB levels for intervention around the medium risk line. It is an option to intervene at lower TSB levels for infants closer to 35 wks and at higher TSB levels for those closer to 37 6/7 wk.' This fine adjustment is NOT reflected in the stored figures (medium line only).
Exchange transfusion, Fig.4 footnote verbatim: 'Immediate exchange transfusion is recommended if infant shows signs of acute bilirubin encephalopathy ... or if TSB is >= 5 mg/dL (85 umol/L) above these levels.' This additional safety margin (+5 mg/dL with signs of kernicterus) is not included in the stored exchange values (these are the base thresholds, not a surcharge).
Source PDF for visual verification of the original figures: https://emergency.med.ufl.edu/files/2013/02/management-of-hyperbilirubinemia.pdf (23 pages, headers and footers identical to the official Pediatrics 2004 publication, incl. DOI 10.1542/peds.114.1.297 and page numbers 297-316; publications.aap.org itself was blocked for automated requests (403); no freely accessible full text of this 2004 article was found via PMC).

Found an error or have an idea?

Tell us if something does not work or could be better.

Give feedback

Related calculators

Note

Calculation aid for healthcare professionals. Percentiles and z-scores describe the position of a measured value in the reference population; clinical assessment remains with the treating physician.