Ultrasound Obstet Gynecol 2010; 35: 442–448
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.7605
Incorporation of femur length leads to underestimation
of fetal weight in asymmetric preterm growth restriction
L. K. PROCTOR*, V. RUSHWORTH*, P. S. SHAH†, J. KEUNEN*, R. WINDRIM*, G. RYAN*
and J. KINGDOM*
*Maternal-Fetal Medicine Division, Department of Obstetrics & Gynecology and †Department of Pediatrics (PS), Mount Sinai Hospital,
University of Toronto, Ontario, Canada
K E Y W O R D S: asymmetric; biometry; birth weight; femur length; intrauterine growth restriction; IUGR; placenta; ultrasound
ABSTRACT
Objective To review the performance of a variety of
biometry formulae for estimated fetal weight (EFW) in
the management of severely growth restricted fetuses with
abnormal umbilical artery Doppler at a single perinatal
institution.
Methods Forty-three pregnancies were retrospectively
reviewed. Inclusion criteria were: chromosomally/
structurally normal fetus; complete ultrasound biometry at ≤ 7 days from delivery; EFW < 10th centile;
absent/reversed end-diastolic flow in the umbilical arteries; and delivery at < 32 + 6 weeks. EFW accuracy and
precision were compared among nine formulae utilizing combinations of head circumference (HC), biparietal diameter (BPD), abdominal circumference (AC) and
femur length (FL) measurements.
Results Twenty-six (60.5%) fetuses showed asymmetric growth (HC/AC ratio > 95th centile). Analysis of
the systematic and random errors associated with each
formula showed that the birth weight of asymmetricallygrown fetuses was most closely approximated by the
Hadlock equation that utilized BPD and AC measurements only. The birth weight of symmetrically-grown
fetuses was most closely approximated by EFW derived
from Hadlock equations that utilized ≥ three biometry measurements, including FL. Incorporation of FL
into Hadlock formulae led to significant underestimation
of birth weight in the fetuses with asymmetric growth
(mean percentage error ± SD: EFWFL – AC , −13.3 ±
9.8%; EFWBPD – FL – AC , −10.8 ± 9.8%; EFWHC – FL – AC ,
−11.8 ± 9.3%; EFWBPD – HC – FL – AC , −11.7 ± 9.5%; P <
0.001). The same equations were accurate in fetuses
with symmetric growth (EFWFL – AC , 3.1 ± 10.0%;
EFWBPD – FL – AC , 1.0 ± 8.9%; EFWHC – FL – AC , 0.3 ±
8.7%; EFWBPD – HC – FL – AC , 0.4 ± 15.5%). Use of the best
performing equation (Hadlock 3), which does not include
FL, to estimate weight in asymmetrically-grown fetuses
over 28 weeks’ gestation, would have reduced the proportion of those with an underestimation of fetal weight
of > 100 g from nine (50.0%) to three (16.7%).
Conclusions Biometry methods that exclude FL should
be considered in asymmetric intrauterine growth restriction associated with abnormal umbilical artery Doppler
waveforms. Copyright  2010 ISUOG. Published by
John Wiley & Sons, Ltd.
INTRODUCTION
Intrauterine growth restriction (IUGR) increases the
risks for a range of adverse outcomes including
stillbirth, iatrogenic preterm birth, neonatal death
and neurosensory disability among survivors1 – 3 . These
adverse outcomes are inversely related to gestational age
and are compounded by the underlying diagnosis and
the severity of the growth restriction process2 . Though a
range of diseases may lead to IUGR, the most common
pathology is placental vascular insufficiency4 . The more
severe forms of this disease result in birth before 32 weeks’
gestation and are characterized by abnormal uterine
and umbilical artery Doppler waveforms and a small
placenta5 . Such growth-restricted fetuses are at risk of
stillbirth, which may be prevented via intensive fetal
monitoring and timely Cesarean delivery2 .
The decision to undertake Cesarean delivery in severe
IUGR is based upon several factors including the
biophysical profile score, umbilical artery and fetal
Doppler waveforms and a non-stress test6 . When the
gestational age is > 29 weeks and the birth weight
Correspondence to: Dr J. Kingdom, Department of Obstetrics & Gynecology, Mount Sinai Hospital, 600 University Avenue, Room 3265,
Toronto, ON, Canada M5G 1X5 (e-mail: [email protected])
Accepted: 14 August 2009
Copyright  2010 ISUOG. Published by John Wiley & Sons, Ltd.
ORIGINAL PAPER
IUGR and birth weight
is > 800 g, reported outcomes are generally favorable
and justify Cesarean delivery2 . However, below this
gestational age and especially with birth weights
of < 600 g, the risk of neonatal death and severe
neurodevelopmental disabilities progressively increases;
some women may decline Cesarean delivery in the fetal
interest in favor of induction of labor leading to a nonmonitored intrapartum stillbirth2 .
Of the two key variables, namely gestational age
and birth weight, the precision of gestational age is
usually known to within 7 days since most women
will have had a dating ultrasound examination before
16 weeks’ gestation7 . Potentially much greater errors
exist when estimating fetal weight. The standard
approach is to derive estimated fetal weight (EFW)
from formulae that include varying combinations of
abdominal circumference (AC), femur length (FL)
and head measurements (fetal head circumference
(HC) and biparietal diameter (BPD))8 . Based on the
work of Hadlock et al.9 and Shepard et al.10 , current
practice is to derive EFW from at least two biometry
measurements to minimize errors. Many EFW studies
have not focused specifically on small-for-gestational
age (SGA) fetuses, but those that have addressed this
group found that EFW errors were increased11 – 13 .
In one systematic review of methods for estimating
fetal weight in low-birth-weight fetuses large random
errors were reported, and no specific formula was
recommended13 .
In the 25 years since the publication of multiparameter fetal biometry methods9,10 , ultrasound techniques,
especially Doppler ultrasound, have evolved considerably for the diagnosis and management of severe preterm
IUGR due to placental insufficiency2 . Fetuses with severe
IUGR due to placental vascular insufficiency typically
have abnormal umbilical artery Doppler2 and some
develop asymmetric body proportions described as ‘head
sparing’8,14 . The potential error in fetal weight estimation introduced by altered body proportions in IUGR
due to placental insufficiency may also be compounded
by the development of oligohydramnios, since the margins of the abdomen become progressively more difficult
to define in comparison with a healthy, constitutionally small fetus surrounded by normal amniotic fluid15 .
Finally, since disproportionately short FL measurements
have been noted as an early feature during the evolution
of severe IUGR16,17 , current multiparameter methods for
deriving EFW may not be appropriate for this subset of
small fetuses.
The purpose of this study was to review the
performance of a variety of biometry formulae for EFW in
the management of severely growth restricted fetuses with
abnormal umbilical artery Doppler at a single perinatal
institution.
METHODS
After obtaining hospital research ethics board approval,
we retrospectively reviewed our database for women
Copyright  2010 ISUOG. Published by John Wiley & Sons, Ltd.
443
whose pregnancies were complicated by placentamediated severe IUGR and who delivered at Mount
Sinai Hospital between January 2001 and July 2008.
Inclusion criteria were: singleton pregnancy; birth at or
before 32 + 6 weeks of gestation; estimated fetal weight
< 10th percentile for sex and gestational age according
to birth-weight data for Canadian infants18 ; evidence
of IUGR due to placental vascular insufficiency (absent
or reversed end-diastolic flow velocity in the umbilical
arteries); full ultrasound biometry within 7 days of
delivery (defined as electronically-archived images of HC,
BPD, AC and FL measurements); and alive at the time
of ultrasound examination. We excluded fetuses with
congenital malformations and those with suspected or
confirmed genetic abnormalities.
Ultrasound measurements were made using two types
of ultrasound machine (ATL500 or iu22) from the same
manufacturer (Philips Medical, Andover, MA, USA) by
sonographers, fellows or staff in our fetal medicine unit
using standard techniques19 . The EFWs, which had been
derived automatically by the ultrasound machines and
manually recorded in the ultrasound report filed in each
patient’s chart (EFWch ), were retrospectively reviewed.
All ultrasound images from eligible pregnancies were
reviewed for this study. Actual birth weight at delivery
and EFWch were compared with nine de-novo calculated
estimates of fetal weight (EFWcalc ) derived using various
components of fetal biometry. These comprised seven
equations developed by Hadlock et al.9,20 , one by Shepard
et al.10 and one by Sabbagha et al.21 (Table 1).
Clinical antenatal records and delivery information
were reviewed according to previously published standard clinical outcome criteria5 . Hypertensive disorders
were as defined by the American College of Obstetricians
and Gynecologists’ guideline criteria22 . A significant medical history was defined by one or more of the following:
chronic hypertension; thrombophilia disorder (heterozygote factor V Leiden gene mutation, prothrombin gene
mutation, elevated anticardiolipin antibodies (> 15 GPL),
or presence of lupus anticoagulant); or systemic lupus
erythematosus. A complex obstetric history was defined
by one or more of the following previous outcomes: stillbirth at > 20 weeks of gestation; extreme preterm birth
at < 32 weeks of gestation; severe pre-eclampsia; placental abruption; or reported IUGR. Gestational age was
derived from first-trimester measurements of the fetal
crown–rump length or from biparietal diameter (BPD)
before 16 weeks’ gestation.
Descriptive statistics are presented as median and
interquartile range. Fetuses were considered to have
asymmetric growth when the HC/AC ratio was above the
95th centile23 . Differences between EFWs and actual birth
weights (BW) are presented graphically as mean with SD
of the signed percentage error ((EFW − BW) ×100/BW%).
Mean percentage errors were compared to zero using the
paired t-test. The formulae were compared for systematic
and random errors using the paired t-test and the
correlated variances test24 , respectively25,26 . Differences
in accuracy (systematic errors) between formulae were
Ultrasound Obstet Gynecol 2010; 35: 442–448.
Proctor et al.
444
Table 1 Equations used to derive estimates of fetal weight
Formula
Equation
Hadlock 120
Hadlock 220
Hadlock 320
Hadlock 420
Hadlock 59
Hadlock 69
Hadlock 79
Shepard10
Sabbagha21
Ln(BW) = 2.695 + 0.253(AC) − 0.00275(AC)2
Log10 (BW) = 1.182 + 0.0273(HC) + 0.07057(AC) − 0.00063(AC)2 − 0.0002184(HC)(AC)
Log10 (BW) = 1.1134 + 0.05845(AC) + 0.000604(AC)2 − 0.007365(BPD)2 + 0.000595(BPD)(AC) + 0.1695(BPD)
Log10 (BW) = 1.3598 + 0.051(AC) + 0.1844(FL) − 0.0037(FL)(AC)
Log10 (BW) = 1.335 − 0.0034(AC)(FL) + 0.0316(BPD) + 0.0457(AC) + 0.1623(FL)
Log10 (BW) = 1.326 − 0.00326(AC)(FL) + 0.0107(HC) + 0.0438(AC) + 0.158(FL)
Log10 (BW) = 1.3596 + 0.0064(HC) + 0.0424(AC) + 0.174(FL) + 0.00061(BPD)(AC) − 0.00386(AC)(FL)
Log10 (BW) = −1.7492 + 0.166(BPD) + 0.046(AC) − (2.646(AC + BPD))/1000
BW = 1849.4 − (47.13)(SUM)+(0.37721)(SUM)2, where SUM = GA + HC + 2(AC) + FL
AC, abdominal circumference; BPD, biparietal diameter; BW, birth weight; FL, femur length; GA, gestational age; HC, head circumference.
considered significant when P was less than 0.006,
as determined by Bonferroni correction for multiple
comparisons. Differences in precision (random errors)
were considered significant when r values were greater
than 0.381 (asymmetric IUGR) or 0.482 (symmetric
IUGR) depending on sample size24 . Statistical analysis
was performed using SigmaStat 3.1 software (Systat
Software Inc., Chicago, IL, USA).
RESULTS
Forty-three pregnancies met the inclusion criteria and
their clinical characteristics and outcomes are summarized
in Tables 2 and 3, respectively. All 24 (55.8%) cases in
which there was perinatal death (21 delivered vaginally,
three delivered by Cesarean section) had a weight at
delivery ≤ 10th centile for gestational age and sex.
All 21 (48.8%) vaginal deliveries were inductions for
either intrauterine fetal demise following the EFW
ultrasound (median birth weight, 585 (range, 539–660) g;
n = 4) or for extreme IUGR (median birth weight,
560 (range, 370–835) g; n = 17). The extreme IUGR
group delivering vaginally did not have fetal heart rate
monitoring. These mothers were counseled extensively
by maternal–fetal medicine specialists and neonatologists
regarding predicted outcomes based on gestational age
and EFWch . They elected non-monitored induction of
labor and non-invasive comfort care for the neonate if
born alive.
Analysis of the EFWch showed that the mean (±
SD) percentage error associated with estimating actual
birth weight at delivery was −8.3 ± 11.0% at our
institution. This underestimation differed significantly
from zero (P < 0.001, paired t-test). To investigate
the source of this error we divided the cohort
according to gestational age at delivery and degree
of body asymmetry (Figure 1) and found that the
mean percentage error remained significantly different
from zero in fetuses displaying asymmetric growth
(< 27 weeks, −8.0 ± 6.6%; > 28 weeks, −15.2 ± 8.2%)
but not symmetric growth (< 27 weeks, 0.2 ± 11.8%;
> 28 weeks, −3.8 ± 10.1%) regardless of gestational age.
This suggests that asymmetry is an important source of
Copyright  2010 ISUOG. Published by John Wiley & Sons, Ltd.
Table 2 Clinical characteristics of the cohort (n = 43)
Characteristic
Median (IQR) or n (%)
Age (years)
Primigravid
Smoker
Ethnicity
Caucasian
Asian
African
Complex obstetric or medical history
Complex obstetric history
Stillbirth at > 20 weeks
Pre-eclampsia/HELLP
Previous preterm delivery
Recurrent miscarriage (≥ 3)
Abruption
Previous IUGR infant
Significant medical history
Autoimmune disorder
Thrombophilia (n = 21 tested)
Chronic hypertension
33 (29–36)
16 (37.2)
1 (2.3)
27 (62.8)
13 (30.2)
3 (7.0)
22 (51.2)
9 (20.9)
2 (4.7)
3 (7.0)
5 (11.6)
3 (7.0)
0 (0)
1 (2.3)
17 (39.5)
1 (2.3)
2 (4.7)
14 (32.6)
IQR, interquartile range; IUGR, intrauterine growth restriction.
error when estimating fetal weight in severe early-onset
IUGR.
De-novo estimations of fetal weight were derived using
fetal biometry measurements (HC, BPD, AC, FL) and
the seven Hadlock equations, the Shepard equation
and the Sabbagha equation (Table 1 and Figure 2).
Initial examination of the accuracy of each formula
in comparison with the birth weight showed that only
two formulae (Hadlock 3 and Shepard) showed a mean
percentage error within 5% of zero in the asymmetric
subgroup, and the mean error did not differ significantly
from zero for either of these (Figure 2a, paired t-test).
Interestingly, these formulae incorporated measurements
of BPD and AC only. In contrast, most of the formulae
showed a mean percentage error within 5% of zero
in the symmetric subgroup, with the mean error only
differing significantly from zero for the Sabbagha formula
(Figure 2b, paired t-test). The mean percentage error in
this group was the lowest with Hadlock formulae 5–7,
which utilized three or more biometric measurements,
including FL (Figure 2b).
Ultrasound Obstet Gynecol 2010; 35: 442–448.
IUGR and birth weight
445
Table 3 Clinical outcomes of the cohort (n = 43)
Outcome
Median (range) or n (%)
Cesarean delivery
Male : female ratio
Alive
Stillbirth/IUFD/TOP
Neonatal death
Gestational age at delivery*
< 24 weeks
24 to 27 + 6 weeks
28 to 32 + 6 weeks
Birth weight (g)
Weight centile at delivery
< 5th centile
5th –10th centile
> 10th centile
Pre-eclampsia/HELLP
22 (51.2)
28 : 15 (65.1 : 34.9)
19 (44.2)
20 (46.5)
4 (9.3)
28 (23–32)
2 (4.7)
18 (41.9)
23 (53.5)
660 (370–1000)
32 (74.4)
8 (18.6)
3 (7.0)
25 (58.1)
*Gestational age was derived from first-trimester measurements of
fetal crown–rump length, or from biparietal diameter before 16
weeks’ gestation. IUFD, intrauterine fetal demise; TOP, termination
of pregnancy.
that the Hadlock 3 formula is the best predictor of fetal
weight in asymmetrically-grown fetuses.
Of the six other equations with mean percentage
error within 5% of zero, none had a significantly
different systematic error from the Hadlock 5 formula
in the symmetric IUGR subgroup (Table 4, paired
t-test). However, the random errors of the Hadlock 3
and Shepard formulae were significantly greater (Table 4,
correlated variances test). This suggests that while
formulae that incorporate BPD and AC alone better
predict birth weight in asymmetric IUGR, the inclusion
of FL increases the accuracy and precision when the fetus
displays symmetric growth.
Utilizing the Hadlock formula that incorporates BPD
and AC alone (Hadlock 3) to estimate fetal weight in
asymmetrically-grown fetuses over 28 weeks of gestation
would have reduced the proportion of fetuses with an
underestimation of fetal weight of > 100 g from nine
(50%) using EFWch , to three (17%).
DISCUSSION
30
Percentage error
20
10
0
−10
−20
−30
Asym.
Sym.
Asym.
Sym.
23 to 27 weeks
28 to 32 weeks
Gestational age
Figure 1 Mean and SD of the percentage error, in comparison to
birth weight, of ultrasound estimations of fetal weight recorded in
patient charts according to gestational age and body proportion.
Values below 0 denote underestimation of birth weight.
*Significant difference from zero (paired t-test). Dotted lines
indicate ± 5% error. Asym., asymmetric; Sym., symmetric.
Based on these findings we evaluated the accuracy
(systematic error) and precision (random error) of each
formula in comparison with the Hadlock 3 and 5
formulae in asymmetric and symmetric IUGR, respectively
(Table 4). The Hadlock 5 formula (incorporating BPD,
AC and FL measurements) was chosen, despite the lack
of evidence that it is the most accurate predictor of birth
weight in symmetric IUGR, to emphasize the effect of the
addition of FL in comparison with the Hadlock 3 formula
(incorporating BPD and AC measurements only). In the
asymmetric subgroup, only the accuracy and precision
of the Sabbagha formula did not differ significantly from
those of the Hadlock 3 formula (Table 4, paired t-test and
correlated variances test). However, the mean percentage
error of the Sabbagha formula did exceed 5%, suggesting
Copyright  2010 ISUOG. Published by John Wiley & Sons, Ltd.
Despite advances in neonatal intensive care, the postnatal
progress of extremely growth-restricted fetuses born with
a weight below 600 g remains guarded, with high rates of
neonatal death and long-term handicap1,3 . The important
task of counseling women in the prenatal period has
been made easier via the emergence of pooled data from
various neonatal networks1,2 . Of the two basic criteria
required to counsel women with a severely IUGR fetus,
namely gestational age and fetal weight, the error in
the determination of gestational age in contemporary
obstetrics is usually trivial because most women in such
circumstances have undertaken an integrated prenatal
screening test where the error in ultrasound dating is fewer
than 7 days7 . By contrast, fetal weight determination is
subject to greater errors8 , to the extent that the term
‘estimated fetal weight’ pervades ultrasound reporting, in
contrast to the more decisive term ‘gestational age’.
Current American27 and British28 clinical practice
guidelines for the management of IUGR differ in their
recommendations regarding the optimal method of biometry to be used for estimating fetal weight in IUGR29 .
Though the American College of Obstetricians and Gynecologists does not recommend a specific formula, Hadlock
et al.20 and Shepard et al.10 are both cited in their guidelines, suggesting that multiple parameters may be used
to derive EFW via various equations27 . By contrast the
more recent RCOG guideline discusses the methodological quality of various studies30 – 34 , suggesting that the
Shepard10 formula (incorporating AC and BPD) or the
Aoki35 formula (incorporating AC, BPD and FL) resulted
in the fewest errors. These studies were not derived solely
from newborns with birth weights under 1000 g, and thus
cannot be generalized to extreme IUGR fetuses12 .
With progressively more extreme degrees of IUGR,
resulting in iatrogenic birth under 1000 g by Cesarean
section, sonographers and perinatal obstetricians have
sought additional confirmatory evidence of IUGR. These
Ultrasound Obstet Gynecol 2010; 35: 442–448.
Proctor et al.
446
30
(b)
20
EFW formula
Sabbagha
Shepard
Hadlock 7
Sabbagha
Shepard
Hadlock 7
Hadlock 6
Hadlock 5
Hadlock 4
−30
Hadlock 3
−30
Hadlock 2
−20
Hadlock 1
−20
Hadlock 6
−10
Hadlock 5
−10
0
Hadlock 4
0
10
Hadlock 3
10
Hadlock 2
Percentage error
20
Percentage error
30
Hadlock 1
(a)
EFW formula
Figure 2 Mean and SD of the percentage error of de-novo ultrasound estimates of fetal weight, using formulae published by Hadlock et al.
(1–7)9,20 , Shepard et al.10 and Sabbagha et al.21 , for the subset of intrauterine growth-restricted fetuses with asymmetric body proportions
(n = 26) (a) and the subset with symmetric body proportions (n = 17) (b). Values below 0 denote underestimation of birth weight.
*Significant differences from zero (paired t-test). Dotted lines indicate ± 5% error. EFW, estimated fetal weight.
Table 4 Accuracy and precision of ultrasound estimations of fetal weight in placenta-mediated intrauterine growth restriction (IUGR). The
percent error (PE) in estimating fetal weight, in comparison to birth weight, is shown for each formula according to body proportion.
Comparisons of the systematic error and the random error are shown for each of the formulae in comparison with the Hadlock 3 formula in
the asymmetric IUGR group and the Hadlock 5 formula in the symmetric IUGR group
Asymmetric IUGR (n = 26)
Symmetric IUGR (n = 17)
Formula
PE ± SD (%)
P
r
PE ± SD (%)
P
r
Hadlock 120
Hadlock 220
Hadlock 320
Hadlock 420
Hadlock 59
Hadlock 69
Hadlock 79
Shepard10
Sabbagha21
−12.5 ± 10.8
−7.5 ± 9.7
−1.4 ± 11.2
−13.3 ± 9.8
−10.8 ± 9.8
−11.8 ± 9.3
−11.7 ± 9.5
2.3 ± 14.3
5.1 ± 14.1
< 0.001
< 0.001
—
<0.001
< 0.001
< 0.001
< 0.001
0.001
0.018
0.054
0.232
—
0.163
0.196
0.243
0.224
0.618
0.249
9.4 ± 14.9
2.9 ± 10.6
4.8 ± 15.8
3.1 ± 10.0
1.0 ± 8.9
0.3 ± 8.7
0.4 ± 15.5
3.6 ± 15.5
9.4 ± 11.0
0.008
0.304
0.212
0.079
—
0.477
0.334
0.412
0.004
0.574
0.231
0.619
0.221
—
0.072
0.148
0.575
0.230
Differences between systematic (mean percent) errors were considered significant when P was < 0.006 using the paired t-test. Differences
between random errors (SDs) were considered significant when r values were greater than 0.381 (asymmetric) or 0.482 (symmetric) using the
correlated variances test.
additional features include ‘head-sparing’ with an elevated
HC/AC ratio, oligohydramnios and abnormal umbilical
artery Doppler waveforms2 . Our data show that the
type of IUGR, symmetric vs. asymmetric, influences the
errors produced by the different biometry methods. In
asymmetric severely IUGR fetuses, the incorporation
of FL resulted in a significant underestimation of fetal
weight. These findings may seem in contrast to those of
Hadlock et al. (1985), who concluded that the accuracy
of ultrasound estimations of fetal weight improves
with additional biometry measurements9 . We reconcile
this difference by noting that Hadlock et al. developed
their formulae on fetuses with normal growth patterns.
Nevertheless, our observations are consistent with recent
studies showing that disproportionately short femurs
are an early feature of severe IUGR16,17 . We therefore
Copyright  2010 ISUOG. Published by John Wiley & Sons, Ltd.
recommend that FL be excluded in the estimation
of fetal weight in asymmetric IUGR fetuses. In such
circumstances, the use of a Hadlock equation based on
BPD and AC measurements will result in acceptable mean
percentage errors below 5%. In our cohort, this change in
practice would have reduced the number of fetuses with
asymmetric growth delivering at more than 28 weeks’
gestation with an underestimation of fetal weight of
> 100 g from nine of 18 (50.0%) to three (16.7%).
By contrast, when the fetus is symmetrically small, the
conventional approach to include measurements of all
three body structures seems reasonable, since the Hadlock
equations that include these data consistently produced
low percentage errors.
The concept of employing a diagnosis-specific method
of weighting individual components of biometry was
Ultrasound Obstet Gynecol 2010; 35: 442–448.
IUGR and birth weight
proposed by Sabbagha et al. in 198921 . Their formula
differs from the others in that it was developed for the SGA
fetus and in addition to HC, AC and FL measurements
it includes gestational age. In their study they found
a 9.3% absolute error in a subset of nine preterm
SGA fetuses. We found that the Sabbagha formula had
similarly low random and systematic errors, compared
with the Hadlock 3 formula, when estimating birth weight
in asymmetric IUGR fetuses. However, since the mean
percentage error of the Sabbagha formula exceeded 5%
and because many machine software systems currently
utilize Hadlock formulae, the pragmatic approach is to
choose a Hadlock formula that does not include FL in
asymmetric IUGR.
There is evidence that the observed error in estimating
fetal weight may be improved through the use of
three-dimensional ultrasound measurements of fetal
soft tissue25 . Siemer et al. showed that volumetric
measurements of the fetal abdomen and femur improved
the accuracy of weight estimation in 150 small fetuses
(< 1600 g)26 . Though this technique has yet to be used in
asymmetric IUGR, it may play a role improving accuracy.
In summary, our data expose an important and
preventable cause of underestimating weight – which,
therefore, affects the prognosis – in asymmetric severely
IUGR fetuses with absent or reversed end-diastolic
flow velocities in the umbilical arteries. Sonographers,
maternal–fetal medicine specialists and neonatologists
should be familiar with the method of biometry they
are using in this context, so as to minimize errors. Further
research is needed to improve the accuracy of fetal weight
determination in extreme IUGR.
ACKNOWLEDGMENTS
The authors acknowledge funding from Rose Torno
Chair in Obstetrics & Gynecology, Mount Sinai Hospital
(to J. K.).
REFERENCES
1. Bernstein IM, Horbar JD, Badger GJ, Ohlsson A, Golan A.
Morbidity and mortality among very-low-birth-weight neonates
with intrauterine growth restriction. The Vermont Oxford
Network. Am J Obstet Gynecol 2000; 182: 198–206.
2. Baschat AA, Cosmi E, Bilardo CM, Wolf H, Berg C, Rigano S,
Germer U, Moyano D, Turan S, Hartung J, Bhide A, Müller T,
Bower S, Nicolaides KH, Thilaganathan B, Gembruch U, Ferrazzi E, Hecher K, Galan HL, Harman CR. Predictors of neonatal outcome in early-onset placental dysfunction. Obstet
Gynecol 2007; 109: 253–261.
3. Baschat AA, Viscardi RM, Hussey-Gardner B, Hashmi N, Harman C. Infant neurodevelopment following fetal growth restriction: relationship with antepartum surveillance parameters.
Ultrasound Obstet Gynecol 2009; 33: 44–50.
4. Krebs C, Macara LM, Leiser R, Bowman AW, Greer IA, Kingdom JC. Intrauterine growth restriction with absent enddiastolic flow velocity in the umbilical artery is associated with
maldevelopment of the placental terminal villous tree. Am J
Obstet Gynecol 1996; 175: 1534–1542.
Copyright  2010 ISUOG. Published by John Wiley & Sons, Ltd.
447
5. Toal M, Keating S, Machin G, Dodd J, Adamson SL, Windrim RC, Kingdom JC. Determinants of adverse perinatal outcome in high-risk women with abnormal uterine artery Doppler
images. Am J Obstet Gynecol 2008; 198: 330.e1–e7.
6. Baschat AA. Pathophysiology of fetal growth restriction:
implications for diagnosis and surveillance. Obstet Gynecol
Surv 2004; 59: 617–627.
7. Sheppard C, Platt LD. Nuchal translucency and first trimester
risk assessment: a systematic review. Ultrasound Q 2007; 23:
107–116.
8. Degani S. Fetal biometry: clinical, pathological, and technical
considerations. Obstet Gynecol Surv 2001; 56: 159–167.
9. Hadlock FP, Harrist RB, Sharman RS, 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: 333–337.
10. Shepard MJ, Richards VA, Berkowitz RL, Warsof SL, Hobbins JC. An evaluation of two equations for predicting fetal
weight by ultrasound. Am J Obstet Gynecol 1982; 142: 47–54.
11. Platz E, Newman R. Diagnosis of IUGR: traditional biometry.
Semin Perinatol 2008; 32: 140–147.
12. Smith GC, Smith MF, McNay MB, Fleming JE. The relation
between fetal abdominal circumference and birthweight:
findings in 3512 pregnancies. Br J Obstet Gynaecol 1997;
104: 186–190.
13. Dudley NJ. A systematic review of the ultrasound estimation of
fetal weight. Ultrasound Obstet Gynecol 2005; 25: 80–89.
14. Snijders RJ, Sherrod C, Gosden CM, Nicolaides KH. Fetal
growth retardation: associated malformations and chromosomal abnormalities. Am J Obstet Gynecol 1993; 168: 547–555.
15. Divon MY, Chamberlain PF, Sipos L, Platt LD. Underestimation of fetal weight in premature rupture of membranes.
J Ultrasound Med 1984; 3: 529–531.
16. Weisz B, David AL, Chitty L, Peebles D, Pandya P, Patel P,
Rodeck CH. Association of isolated short femur in the midtrimester fetus with perinatal outcome. Ultrasound Obstet
Gynecol 2008; 31: 512–516.
17. Papageorghiou AT, Fratelli N, Leslie K, Bhide A, Thilaganathan B. Outcome of fetuses with antenatally diagnosed
short femur. Ultrasound Obstet Gynecol 2008; 31: 507–511.
18. Kramer MS, Platt RW, Wen SW, Joseph KS, Allen A, Abrahamowicz M, Blondel B, Bréart G; Fetal/Infant Health Study
Group of the Canadian Perinatal Surveillance System. A new
and improved population-based Canadian reference for birth
weight for gestational age. Pediatrics 2001; 108: E35.
19. Chudleigh T, Thilaganathan B. Obstetric Ultrasound: How,
Why and When. In Routine second trimester screening – assessing gestational age. Churchill Livingstone: Edinburgh; London; New York; Oxford; Philadelphia; St Louis;
Sydney; Toronto, 2004.
20. Hadlock FP, Deter RL, Harrist RB, Park SK. Estimating fetal
age: computer-assisted analysis of multiple fetal growth
parameters. Radiology 1984; 152: 497–501.
21. Sabbagha RE, Minogue J, Tamura RK, Hungerford SA. Estimation of birth weight by use of ultrasonographic formulas
targeted to large-, appropriate-, and small-for-gestational-age
fetuses. Am J Obstet Gynecol 1989; 160: 854–860; discussion
860–862.
22. ACOG Committee on Practice Bulletins – Obstetrics. ACOG
practice bulletin. Diagnosis and management of preeclampsia
and eclampsia. Number 33, January 2002. Obstet Gynecol
2002; 99: 159–167.
23. Snijders RJ, Nicolaides KH. Fetal biometry at 14–40 weeks’
gestation. Ultrasound Obstet Gynecol 1994; 4: 34–48.
24. Snedecor GW, Cochran WG. Statistical Methods. Comparison
of two correlated variances in paired samples. Iowa State
University Press: Ames, 1980.
25. Lee W, Deter RL, Ebersole JD, Huang R, Blanckaert K,
Romero R. Birth weight prediction by three-dimensional ultrasonography: fractional limb volume. J Ultrasound Med 2001;
20: 1283–1292.
Ultrasound Obstet Gynecol 2010; 35: 442–448.
448
26. Siemer J, Hilbert A, Hart N, Hoopmann M, Schneider U,
Girschick G, Müller A, Schild RL. Specific weight formula for
fetuses with abdominal wall defects. Ultrasound Obstet Gynecol
2008; 31: 397–400.
27. ACOG Committee on Practice Bulletins – Gynecology.
Intrauterine growth restriction. Clinical management guidelines
for obstetrician–gynecologists. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet 2001; 72:
85–96.
28. Royal College of Obstetricians and Gynaecologists (RCOG).
The Investigation and Managment of the Small-for-GestationalAge Fetus. Clinical Guidelines No. 31. RCOG Press: London,
2002.
29. Chauhan SP, Gupta LM, Hendrix NW, Berghella V; American
College of Obstetricians and Gynecologists. Intrauterine growth
restriction: comparison of American College of Obstetricians
and Gynecologists practice bulletin with other national
guidelines. Am J Obstet Gynecol 2009; 200: e.1–6.
Copyright  2010 ISUOG. Published by John Wiley & Sons, Ltd.
Proctor et al.
30. Deter RL, Harrist RB, Hadlock FP, Carpenter RJ. The use of
ultrasound in the assessment of normal fetal growth: a review.
J Clin Ultrasound 1981; 9: 481–493.
31. Chang TC, Robson SC, Boys RJ, Spencer JA. Prediction of the
small for gestational age infant: which ultrasonic measurement
is best? Obstet Gynecol 1992; 80: 1030–1038.
32. Deter RL, Hadlock FP, Harrist RB, Carpenter RJ. Evaluation
of three methods for obtaining fetal weight estimates using
dynamic image ultrasound. J Clin Ultrasound 1981; 9:
421–425.
33. Hill LM, Breckle R, Wolfgram KR, O’Brien PC. Evaluation of
three methods for estimating fetal weight. J Clin Ultrasound
1986; 14: 171–178.
34. Chien PF, Owen P, Khan KS. Validity of ultrasound estimation
of fetal weight. Obstet Gynecol 2000; 95: 856–860.
35. Aoki M. Fetal Weight Calculation; Osaka University Methods.
In Ultrasound in Obstetrics and Gynaecology, Yoshihide C
(ed.). Kinpodo: Kyoto, 1990.
Ultrasound Obstet Gynecol 2010; 35: 442–448.