Highlight
- Significant disparity exists between chart- and equation-derived centiles in the Hadlock fetal weight reference.
- The 3rd and 10th centiles on the published chart underestimate the proportion of small for gestational age (SGA) fetuses compared to the original regression equation.
- Use of the chart results in 3% fewer fetuses classified below the 10th centile and 1.5% fewer below the 3rd centile in clinical populations.
- Equation-derived centiles align more closely with expected population proportions, indicating a need to revisit clinical thresholds and guidelines based on the Hadlock chart.
Background
Accurate estimation of fetal weight (EFW) is pivotal in obstetric care, guiding the diagnosis and management of fetal growth abnormalities, particularly small for gestational age (SGA) fetuses, which are associated with increased perinatal morbidity and mortality. The Hadlock reference, established in 1984 and widely adopted internationally, provides normative ranges and centiles for EFW based on fetal biometrics obtained by ultrasound. However, clinical application depends heavily on the integrity and internal consistency of these reference centiles, which influence thresholds defining fetal growth restriction and recruitment for interventions such as intensified surveillance or early delivery.
Recent clinical observations have suggested potential discrepancies between the published Hadlock chart and the underlying regression equation from which the reference was derived, raising concerns about possible underdiagnosis of SGA fetuses. Given the widespread reliance on Hadlock references in fetal growth assessment, an in-depth methodological evaluation is warranted to safeguard clinical decision-making.
Study Design
This study undertook a methodological reanalysis of the original Hadlock EFW reference using advanced statistical techniques including functional data analysis (FDA) and penalised functional regression within a generalized additive model framework (PFFR-GAM). This enabled a rigorous assessment of the internal consistency between the originally published chart-derived centiles and those computed directly from the regression equation.
Additionally, a retrospective observational audit was performed on 21,874 women with singleton, non-anomalous pregnancies undergoing growth scans between 35+0 and 36+6 weeks of gestation at a tertiary maternity hospital in Oxford, UK. For each fetus, EFW centiles were calculated using both the published chart and the original regression equation to compare classification rates below clinically critical 3rd and 10th centiles.
Key Findings
The analysis revealed a statistically significant systematic disparity between chart-derived and equation-derived centiles. Specifically, the published Hadlock chart’s 3rd and 10th centiles corresponded approximately to the 1st and 6th centiles, respectively, when derived from the regression equation, indicating notable underestimation of smaller fetuses by the chart.
Conversely, the equation-derived 10th centile corresponded approximately to the 17th centile on the chart, further illustrating inconsistency.
In the clinical cohort, using the published chart, 4.2% of fetuses were classified below the 10th centile, whereas the equation-derived centiles identified 7.2% below this threshold, amounting to an absolute underestimation of 3.0% (95% CI 2.8%-3.3%; p<0.001). Similarly, the 3rd centile classification found 0.8% fetuses by chart versus 2.3% by equation-derived centiles, with an absolute difference of 1.5% (95% CI 1.3%-1.7%; p<0.001).
These findings were robust across the gestational age window studied. The use of the regression equation also yielded centiles that matched closer to the nominal expected proportions within the studied population, fulfilling statistical expectations absent in the chart-derived centiles.
This misclassification has significant clinical implications: fewer fetuses at risk of growth restriction and potential adverse outcomes might be identified using the chart, potentially impacting prenatal surveillance strategies and timing of delivery.
Expert Commentary
The Hadlock reference has been a cornerstone in fetal growth assessment for decades. However, this study underscores limitations related to how reference centiles were originally presented and subsequently adopted into clinical practice. A potential source of error includes the digitization and interpolation methods used to create published charts, which may have introduced systematic biases.
Modern statistical approaches like FDA and penalized regression provide more accurate modeling of growth trajectories and better reflect the underlying biological variability.
Clinicians and guideline developers should be cautious relying exclusively on published EFW charts without verifying their alignment with original equations or population-specific data. This is especially pertinent when decisions hinge on strict centile cut-offs for SGA.
Limitations of this study include its retrospective design and relatively narrow gestational age window in the clinical audit, which may not capture all clinical scenarios. Nonetheless, the large sample size and rigorous analytic methods lend weight to the findings.
Future studies should replicate these analyses across different populations and gestational ages and evaluate the clinical outcomes related to reclassification of SGA using updated centiles.
Conclusion
This methodological analysis reveals a critical internal inconsistency within the widely used Hadlock fetal weight reference. The discrepancy between published chart- and equation-derived centiles leads to underestimation of SGA fetuses by the chart, which may contribute to missed diagnoses of fetal growth restriction.
Adoption of the equation-derived centiles provides a closer approximation to expected population proportions and could enhance clinical identification of at-risk pregnancies. Obstetric practice should consider recalibrating growth assessment protocols incorporating these findings to improve fetal surveillance and neonatal outcomes.
Revalidation of fetal weight references in contemporary populations using robust statistical methods is essential to ensure accurate and effective clinical care.
Funding and Registrations
No specific funding was reported for this study. The study was conducted at the Oxford tertiary maternity hospital under institutional approvals.
References
1. Hadlock FP, Harrist RB, Carpenter RJ, Deter RL, Park SK. Estimation of fetal weight with the use of head, body, and femur measurements–a prospective study. Am J Obstet Gynecol. 1985;151(3):333-7.
2. Ioannou C, Papageorghiou AT. Methodological considerations in fetal weight estimation. Ultrasound Obstet Gynecol. 2020;55(5):702-710.
3. Figueras F, Gratacos E. Update on the diagnosis and classification of fetal growth restriction and proposal of a stage-based management protocol. Fetal Diagn Ther. 2014;36(2):86-98.
4. Gardosi J, Mongelli M, Wilcox M, Chang A. An adjustable fetal weight standard. Ultrasound Obstet Gynecol. 1995;6(3):168-74.
5. Ioannou C, Adu-Bredu TK, Twumasi C, Impey L, Mathewlynn S. Internal Consistency of the Hadlock Estimated Fetal Weight Reference: Methodological Analysis. BJOG. 2026 Jun 30; PMID: 42381153.
