Ultrasound Obstet Gynecol 2018; 52: 430–441

Published online 6 September 2018 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/uog.19117

Predictive accuracy of cerebroplacental ratio for adverse

perinatal and neurodevelopmental outcomes in suspected
fetal growth restriction: systematic review and meta-analysis
A. CONDE-AGUDELO1 , J. VILLAR2,3 , S. H. KENNEDY2,3 and A. T. PAPAGEORGHIOU2,3
1 Perinatology Research Branch, Eunice Kennedy Shriver National Institute of Child Health and Human Development/National Institutes of

Health/Department of Health and Human Services, Bethesda, MD, and Detroit, MI, USA; 2 Nuffield Department of Obstetrics &
Gynaecology, University of Oxford, Women’s Centre, John Radcliffe Hospital, Oxford, UK; 3 Oxford Maternal & Perinatal Health
Institute, Green Templeton College, Oxford, UK

K E Y W O R D S: cerebroplacental ratio; Doppler; fetal growth restriction; neurodevelopmental outcomes; perinatal outcome;
predictive accuracy

ABSTRACT ranging from 1.1 to 2.5 and 0.3 to 0.9, respectively.

Objective The cerebroplacental ratio (CPR) has been
proposed for the routine surveillance of pregnancies
with suspected fetal growth restriction (FGR), but the
predictive performance of this test is unclear. The aim
of this study was to determine the accuracy of CPR
for predicting adverse perinatal and neurodevelopmental
outcomes in suspected FGR.
Methods PubMed, EMBASE, CINAHL and Lilacs
were searched from inception to 31 July 2017 for
cohort or cross-sectional studies reporting on the
accuracy of CPR for predicting adverse perinatal and/or
neurodevelopmental outcomes in singleton pregnancies
with FGR suspected antenatally based on sonographic
parameters. Summary receiver–operating characteristics
(ROC) curves, pooled sensitivities and specificities, and
summary likelihood ratios (LRs) were generated.
Results Twenty-two studies (including 4301 women) met
the inclusion criteria. Summary ROC curves showed that
the best predictive accuracy of CPR was for perinatal
death and the worst was for neonatal acidosis, with
areas under the summary ROC curves of 0.83 and
0.57, respectively. The predictive accuracy of CPR was
moderate to high for perinatal death (pooled sensitivity
and specificity of 93% and 76%, respectively, and
summary positive and negative LRs of 3.9 and 0.09,
respectively) and low for composite of adverse perinatal
outcomes, Cesarean section for non-reassuring fetal
status, 5-min Apgar score < 7, admission to the neonatal
intensive care unit, neonatal acidosis and neonatal
morbidity, with summary positive and negative LRs
An abnormal CPR result had moderate accuracy for
predicting small-for-gestational age at birth (summary
positive LR of 7.4). CPR had a higher predictive accuracy
in pregnancies with suspected early-onset FGR. No study
provided data for assessing the predictive accuracy of
CPR for adverse neurodevelopmental outcome.
Conclusion CPR appears to be useful in predicting
perinatal death in pregnancies with suspected FGR.
Nevertheless, before incorporating CPR into the routine
clinical management of suspected FGR, randomized
controlled trials should assess whether the use of
CPR reduces perinatal death or other adverse perinatal
outcomes. Copyright © 2018 ISUOG. Published by John
Wiley & Sons Ltd.
INTRODUCTION
Fetal growth restriction (FGR) is a major clinical
and public health challenge around the world1,2 .
Small-for-gestational age (SGA) at birth, based on
different cut-off values, is a commonly used proxy
measure of FGR3 . FGR is associated with an increased risk
of short- and long-term morbidity and mortality, as well
as impaired neurological and cognitive development4–11 .
Suspected FGR is defined in the antenatal period by
sonographic estimation of fetal anthropometric measures
using a wide range of seldom validated definitions and
cut-off values12–16 . The clinical management of suspected
FGR is challenging and no consensus exists for the best
way to monitor fetal wellbeing in these pregnancies;
consequently, clinical practice varies considerably around

Correspondence to: Prof. J. Villar, Nuffield Department of Obstetrics & Gynaecology, University of Oxford, Women’s Centre, John Radcliffe
Hospital, Oxford OX3 9DU, UK (e-mail: jose.villar@obs-gyn.ox.ac.uk)
Accepted: 23 May 2018

Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd. SYSTEMATIC REVIEW
CPR predicts perinatal death in suspected FGR
the world17–19 . The use of umbilical artery (UA) Doppler
velocimetry in high-risk pregnancies, including those with
suspected FGR, has been shown to be associated with
a significant reduction in perinatal mortality and fewer
Cesarean deliveries and inductions of labor20 .
In 1987, Arbeille et al.21 reported that the cerebropla-
cental ratio (CPR), a measure of cerebral centralization
of fetal blood flow, appeared to be superior to either
fetal middle cerebral artery (MCA) or UA Doppler indices
alone in predicting SGA among women with gestational
hypertension. CPR is calculated by dividing the Doppler
index (pulsatility index (PI), resistance index (RI), or sys-
tolic/diastolic ratio (S/D)) of the MCA by that of the UA.
Physiologically, CPR represents the interaction of alter-
ations in blood flow to the brain, as manifest by increased
diastolic flow as a result of cerebrovascular dilatation
due to hypoxia and increased placental resistance, lead-
ing to decreased diastolic flow in the UA22 . Integrating
CPR into the clinical management of suspected FGR
has been proposed recently22–27 , but the test’s ability
to predict adverse perinatal outcome in this entity has
been questionned28,29 . Hence, we carried out a system-
atic review and meta-analysis to assess the accuracy of
CPR to predict adverse perinatal and neurodevelopmental
outcomes in FGR suspected antenatally.
METHODS
This systematic review was conducted following
a prospectively prepared protocol and reported in
accordance with recommended methods for system-
atic reviews of diagnostic test accuracy30,31 . The
protocol was registered with PROSPERO in March
2016 (CRD42016036488; available from: http://
www.crd.york.ac.uk/PROSPERO/display_record.php?
ID=CRD42016036488).
Literature search
We searched PubMed, EMBASE, CINAHL and Lilacs
from inception to 31 July 2017 using a combination
of keywords and text words related to ‘cerebroplacental
ratio’ and ‘fetal growth restriction’, without language
restrictions (Appendix S1).
Eligibility criteria
We included cohort or cross-sectional studies reporting
on the accuracy of CPR for predicting adverse perinatal
and/or neurodevelopmental outcomes in singleton preg-
nancies with FGR suspected antenatally based on sono-
graphic parameters, and provided the necessary informa-
tion to generate 2 × 2 tables. Studies were excluded if they:
(1) assessed retrospectively the predictive accuracy of CPR
in infants categorized as SGA or FGR based on postnatal
parameters such as birth weight or other anthropometric
measures, and/or placental histopathology; (2) assessed
CPR in a mix of high-risk pregnancies but did not report
Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.
431
results separately for pregnancies with suspected FGR;
(3) assessed CPR in the general population as a screen-
ing tool; (4) were case–control studies without complete
information for cases with suspected FGR, case series or
reports, editorials, comments, reviews or letters without
original data; (5) reported data for CPR only as mean or
median values; (6) did not publish test accuracy estimates,
or sufficient information to calculate them could not be
retrieved.
One reviewer (A.C.-A.) screened titles and abstracts
of all identified citations and selected potentially eligible
studies. Then, these studies were retrieved and assessed by
the same reviewer for inclusion and data extraction, and
a 10% sample of the papers was examined by a second
independent reviewer (J.V.). Disagreements were resolved
through consensus. In cases of duplicate publication, only
the most recent or complete version was included.
Reference standard outcomes
The reference standard outcomes included the following:
perinatal death; any composite of adverse perinatal
outcomes (as defined in the original study and regardless
of its individual components); Cesarean delivery for
fetal distress/non-reassuring fetal status; 5-min Apgar
score < 7; admission to the neonatal intensive care unit
(NICU); neonatal acidosis; neonatal brain lesion; neona-
tal morbidity other than brain lesion; use of mechanical
ventilation; SGA at birth (birth weight < 10th , < 5th or
< 3rd percentile or > 2 SD below the mean adjusted for
gestational age (GA) based on local population values),
and adverse neurodevelopmental outcome (suspected or
diagnosed developmental delay, cerebral palsy, intellec-
tual disability, vision impairment, hearing loss, cognitive
and behavioral impairment, and motor, communication
or learning disorder at any age in childhood).
Assessment of risk of bias
Risk of bias in each included study was evaluated
by at least one investigator using a modified version
of the Quality Assessment of Diagnostic Accuracy
Studies (QUADAS)-2 tool32 . The following domains were
assessed: study design, description of the test, selection of
test cut-off value, blinding of clinicians to CPR results,
inclusion in the analysis of participants recruited into the
study, and use of interventions aimed to prevent adverse
perinatal outcome based on CPR results. Each domain was
scored as ‘low’, ‘high’ or ‘unclear’ risk of bias (Appendix
S2). We did not calculate a summary score estimating the
overall quality of each study because of the well-known
problems associated with such scores33 .
Data extraction
Data were extracted from each article using a specially
designed form for capturing information on study
characteristics (authors, setting, year of publication,
method of recruiting women, design, prospective or
Ultrasound Obstet Gynecol 2018; 52: 430–441.
432
retrospective data collection, blinding to test results, flow
diagram, completeness of follow-up and reporting of
withdrawals, and use of interventions after performing
the test), patient characteristics (inclusion and exclusion
criteria, sample size, and demographic characteristics),
how the test was carried out (GA at testing, frequency
of test, method of performance of test, type of Doppler
and route, site of measurement, plane in which images
were obtained, Doppler index used (PI, RI or S/D), cut-off
value used, and interval from Doppler examination to
delivery), and reference standard outcomes assessed and
their prevalence.
For each study and for all cut-off values defining
an abnormal CPR result, we extracted the number
of true-positive, false-positive, true-negative and false-
negative test results. When predictive accuracy data
were not available, we recalculated them from the
reported results, including scatterplots and bar graphs.
The corresponding authors of primary studies were
contacted to obtain additional information on methods
used and/or relevant unpublished data. Only three authors
supplied additional data.
Data synthesis
Data extracted from each study were used to construct
2 × 2 contingency tables. When any single cell in these
tables contained a zero, we added 0.5 to each cell
to enable calculation of predictive values34 . Sensitivity
and specificity with 95% CIs were calculated separately
for all Doppler indices and cut-off values used, as
well as reference standard outcomes reported. Then,
summary receiver–operating characteristics (ROC) curves
were constructed for each predefined reference standard
outcome using the hierarchal summary ROC model,
regardless of Doppler indices and cut-off values used
to define abnormality35 . Variation in cut-off values across
studies is taken into account by using this model. Pooled
estimates and 95% CIs of sensitivity and specificity were
generated using random-effects bivariate meta-regression
models36 . For studies that reported results for more than
one Doppler index and/or cut-off value, we selected the
most commonly used. We also calculated areas under the
summary ROC curves with their corresponding 95% CIs,
which allowed for comparison of the predictive accuracy
of CPR for different outcomes37 .
Thereafter, summary likelihood ratios (LRs) with 95%
CIs were calculated from the pooled sensitivities and
specificities38 . A guide for the interpretation of LRs
suggests that LRs > 10 for a positive test result and
LRs < 0.1 for a negative test result generate large changes
from pretest to post-test probabilities of disease; LRs
of 5–10 and 0.1–0.2 generate moderate changes in
probability; LRs of 2–5 and 0.2–0.5 generate small (but
sometimes important) changes in probability; and LRs of
1–2 and 0.5–1 generate minimal (and rarely important)
changes in probability39 . Finally, we planned to calculate
the post-test probabilities of the most important reference
standard outcomes by combining summary LRs obtained
Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.
Conde-Agudelo et al.
from meta-analyses for positive and negative test results
and a global prevalence (pretest probability) of these
reference standard outcomes across the studies39 .
Prespecified subgroup analyses were carried out
to assess the predictive accuracy of CPR for any
composite of adverse perinatal outcome according
to GA at diagnosis or delivery (early-onset (< 32
or < 34 weeks) or late-onset (≥ 32 or ≥ 34 weeks), as
defined by the authors), definition of abnormal CPR
result (MCA-PI/UA-PI ≤ 1.08, MCA-PI/UA-PI < 5th
percentile, or MCA-RI/UA-RI < 1 or < 1.05), interval
from CPR measurement to delivery (≤ 7 or > 7 days)
and definition of suspected FGR used (estimated fetal
weight (EFW) < 10th percentile for GA, or EFW < 10th
percentile for GA and abnormal UA Doppler). In addition,
a post-hoc subgroup analysis according to the use (yes
or no) of the CPR results for managing pregnancies with
suspected FGR was performed. We also assessed the effect
of risk of bias of the included studies on the predictive
accuracy of CPR by performing a sensitivity analysis,
including only studies with a low risk of bias in at least
five of the six domains evaluated.
As is common in diagnostic accuracy studies, we
anticipated that there would be substantial between-study
variation in reported pairs of sensitivity and specificity
values. As forest plots, which display both sensitivity
and specificity, depict estimates with associated CIs, it
is possible to discern the presence of high levels of
heterogeneity when there is little overlap in the CIs from
different studies. In order to investigate formally potential
sources of heterogeneity, we used subgroup analysis and
meta-regression by including covariates defined a priori
(Doppler indices and cut-off values used, definition of
suspected FGR used, GA at diagnosis or delivery, interval
from CPR to delivery and study’s risk of bias) in the
bivariate model, which enabled us to assess the effect of
various factors on the predictive accuracy of CPR40,41 . If
there were at least 10 studies included in a meta-analysis,
we planned to assess publication and related biases by
examining the symmetry of funnel plots using Deeks’
test42 . A P-value of < 0.1 for the slope coefficient indicated
significant asymmetry of the funnel plot.
We used SAS version 9.2 (SAS Institute Inc., Cary,
NC, USA) for the analyses and Review Manager 5.3.5
(The Nordic Cochrane Centre, Copenhagen, Denmark)
to generate forest plots and summary ROC curves.
RESULTS
Selection, characteristics and quality of studies
Of 1191 citations identified initially, 22 studies43–64 ,
including a total of 4301 women/fetuses, met the inclusion
criteria (Figure 1). Two studies were performed using the
same cohort60,61 , one reporting results for all cases of
suspected FGR60 and the other for cases of suspected
early-onset FGR61 . We included the results of the latter
study only in the subgroup analysis according to GA at
diagnosis or delivery.
Ultrasound Obstet Gynecol 2018; 52: 430–441.
CPR predicts perinatal death in suspected FGR
Citations identified through Additional citations identified
search of databases (n = 1191) through other sources (n = 0)
Citations screened after duplicates removed (n = 821)
References excluded after screening of
titles/abstracts (n = 669)
Full-text articles assessed for eligibility (n = 152)
Excluded (n = 130)
• No data on predictive accuracy of CPR (n = 28)
• CPR assessed in mix of high-risk pregnancies
without separate data for suspected FGR (n = 24)
• CPR assessed in infants categorized as SGA or
FGR based on postnatal parameters (n = 23)
• Not a test accuracy study (n = 15)
• CPR assessed in unselected pregnancies (n = 13)
• Review, letter, commentary or editorial (n = 13)
• Duplicate publication (n = 4)
• Insufficient data to construct 2 × 2 tables (n = 4)
• Case–control study (n = 3)
• Other reason (n = 3)
Included in qualitative synthesis (n = 22)
Included in meta-analysis (n = 22)
Figure 1 Study selection process. CPR, cerebroplacental ratio;
FGR, fetal growth restriction; SGA, small-for-gestational age.
The main characteristics of the included studies are
displayed in Table 1. All but two studies44,52 were
performed in European or North American countries.
The sample size ranged from 2948 to 88156 (median,
159). The definitions of suspected FGR used in the
studies were as follows: EFW < 10th percentile for GA (11
studies)44,45,47,50,54,56–59,62,64 , EFW < 10th percentile for
GA and/or abdominal circumference (AC) < 5th percentile
(two studies)60,61 , EFW < 10th percentile for GA and
abnormal UA Doppler indices (two studies)53,63 , AC < 5th
percentile for GA and abnormal UA Doppler indices
(two studies)46,49 , AC < 10th percentile for GA (one
study)43 , AC < 10th percentile for GA on at least two
consecutive measurements, 2 weeks apart (one study)55 ,
EFW < 10th percentile for GA with growth rate slower
than normal and abnormal UA Doppler indices (one
study)48 , EFW or AC < 10th percentile for GA and
abnormal UA Doppler indices (one study)52 , and EFW
below the GA-adjusted mean value minus 2 SD (2.3rd
percentile) or a fall of ≥ 10% weight deviation from
the mean weight between two ultrasound examinations
(one study)51 . Ten studies reported results for suspected
late-onset FGR47,50,51,54,56–59,62,63 , four for suspected
early-onset FGR47,56,58,61 and 14 for suspected FGR at all
GAs43–49,52,53,55,56,58,60,64 .
The most common definitions of an abnormal
CPR result were MCA-PI/UA-PI < or ≤ 1.08 (eight
studies43,47,51–53,56,60,61 ) and MCA-PI/UA-PI < 5th percen-
tile for GA (six studies47,50,54,56,58,62 ). The mean or
Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.
433
median interval between CPR measurement and delivery
was < or ∼48 h in five studies46,48,50,51,62 , ≤ 7 days in six
studies49,53,56,57,59,64 , 7–14 days in three studies44,47,58 ,
> 14 days in five studies45,54,55,60,61 , 1–30 days in one
study52 and unreported in two studies43,63 . Most studies
(n = 16) used the last CPR result before delivery in
analyses44,46–55,57–59,62,64 . Sixteen studies reported
that the CPR results were not used to manage the
pregnancies43–45,47,48,50–53,56,58–63 , one reported that
they were used to manage the pregnancies54 and
the remaining five studies did not report on this
issue46,49,55,57,64 . Eleven studies provided data on the
predictive accuracy of CPR for a composite of adverse
perinatal outcomes43,44,47,53,54,56–60,62 , nine for admis-
sion to the NICU43–45,51,56,58–60,62 , seven for Cesarean
delivery for non-reassuring fetal status43,45,50,58–60,62 ,
six for perinatal death45,52,55,56,58,60 and 5-min
Apgar score < 743,44,51,58,59,64 , five each for neonatal
acidosis50,51,59,62,64 and neonatal brain lesions45,48,52,
55,63
, four for neonatal morbidities other than brain
lesions43,45,46,49 , two for SGA at birth43,59 and one for
use of mechanical ventilation55 . No study provided data
on adverse neurodevelopmental outcome.
The risk of bias in each included study is shown
in Figure 2. Eight studies (36%) fulfilled ≥ 5 criteria,
whereas the remaining 14 studies (64%) had ≥ 2
methodological flaws. The most common deficiencies
were related to blinding of clinicians to CPR results and
inclusion in the analyses of participants recruited into the
study.
Predictive accuracy for adverse perinatal outcomes
Summary ROC curves of CPR for predicting adverse
perinatal outcome in pregnancies with suspected FGR are
shown in Figure 3. The best predictive accuracy was for
perinatal death and the worst was for neonatal acidosis,
with areas under the summary ROC curves of 0.83
(95% CI, 0.74–0.92) and 0.57 (95% CI, 0.51–0.63),
respectively. Similar summary ROC curves were obtained
for the prediction of any composite of adverse perinatal
outcomes, Cesarean delivery for non-reassuring fetal
status and admission to the NICU (areas under the
summary ROC curves between 0.71 and 0.74). The
sensitivity and specificity of CPR to predict adverse
perinatal outcome in suspected FGR in the individual
studies are shown in Figure S1.
Table 2 presents the pooled estimates of the predictive
accuracy of CPR for adverse perinatal outcomes. Overall,
CPR showed a moderate-to-high predictive ability for
perinatal death with pooled sensitivity and specificity of
93% and 76%, respectively, and summary positive and
negative LRs of 3.9 and 0.09, respectively (six studies
including 1495 fetuses with 29 perinatal deaths). CPR had
a low predictive performance for any composite of adverse
perinatal outcomes, Cesarean delivery for non-reassuring
fetal status, 5-min Apgar score < 7, admission to the
NICU, neonatal acidosis, neonatal brain lesion, neonatal
morbidity other than brain lesion and use of mechanical
Ultrasound Obstet Gynecol 2018; 52: 430–441.
 Table 1 Main characteristics of studies included in systematic review reporting on predictive accuracy of cerebroplacental ratio (CPR) for adverse perinatal outcome in singleton pregnancy with fetal
434

growth restriction (FGR) suspected antenatally

GA at diagnosis
Definition of of suspected Abnormal test Interval between
Study (country) Design n suspected FGR FGR (weeks) result definition CPR and delivery Reference standard outcome

Gramellini (1992)43 Retro 45 AC < 10th percentile Range, 30–41 MCA-PI/UA-PI Unreported Adverse perinatal outcome (CD for fetal
(Italy) for GA ≤ 1.08 distress or admission to NICU) and its
individual components, 5-min Apgar
score < 7, neonatal complications
(any of: RDS, intracerebral
hemorrhage, seizures, patent ductus
arteriosus, polycythemia)
Makhseed (2000)44 Prosp 70 EFW < 10th percentile Range, 29–42 MCA-RI/UA-RI Mean, 8.6 and 3.8 days Adverse perinatal outcome (any of: CD
(Kuwait) for GA < 1.05 in normal and for fetal distress, 5-min Apgar
abnormal CPR score < 7 or admission to NICU),
groups, respectively 5-min Apgar score < 7, admission to
NICU
Sterne (2001)45 Prosp 53 EFW < 10th percentile Range, MCA-S/D / Mean, 3.9 and CD for fetal distress, NICU admission,
(USA) for GA 22.9–38.9; UA-S/D ≤ 1.0 2.8 weeks in normal RDS requiring intubation, intracranial
mean, 30.3 and abnormal CPR hemorrhage, perinatal death
groups, respectively

Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.

Makh (2003)46 Prosp 70 AC < 5th percentile for Unreported* MCA-PI/UA-PI Median, 6 h; range, Neonatal anemia (hemoglobin
(USA) GA and UA-PI > 2 SD below 1–19 h < 13 g/dL)
> 2 SD above mean mean for GA
for GA
Odibo (2005)47 Retro 183 EFW < 10th percentile < 34 and ≥ 34 MCA-PI/UA-PI Mean, 9.8 days; range, Adverse perinatal outcome (any of:
(USA) for GA < 1.08 and 0–32 days perinatal death, CD for NRFS,
MCA-PI/UA-PI umbilical artery pH < 7.0, 5-min
< 5th percentile Apgar score < 7, grade 3 or 4 IVH,
for GA PVL, RDS)
Jugović (2007)48 Prosp 29 EFW < 10th percentile > 28† MCA-RI/UA-RI < 1.0 < 48 h Neonatal brain lesion (severe
(Croatia) for GA with growth periventricular echodensities and/or
rate slower than moderate/severe intracranial
normal and hemorrhage)
increased UA-RI
for GA
Manogura (2008)49 Prosp 404 AC < 5th percentile for Unreported§ MCA-PI/UA-PI < 1 week NEC
(Multinational‡) GA and increased > 2 SD below
UA Doppler indices mean for GA
Cruz-Martı́nez (2011)50 Prosp 210 EFW < 10th percentile > 37 MCA-PI/UA-PI ∼48 h CD for NRFS, neonatal metabolic
(Spain) for GA < 5th percentile acidosis (UA pH < 7.15 and base
for GA excess > 12 mEq/L)

Continued over
Conde-Agudelo et al.

Ultrasound Obstet Gynecol 2018; 52: 430–441.

 Table 1 Continued

GA at diagnosis
Definition of of suspected Abnormal test Interval between
Study (country) Design n suspected FGR FGR (weeks) result definition CPR and delivery Reference standard outcome
51
Fu (2011) Prosp 126 EFW below GA-adjusted ≥ 36 MCA-PI/UA-PI Median, ≤ 1 day; CD, 5-min Apgar score < 7, umbilical
(Sweden) mean value minus 2 SD < 1.08 range, 0–21 days cord arterial pH < 7.10 at birth,
(2.3rd percentile), or admission to NICU, positive oxytocin
fall of ≥ 10% weight challenge test
deviation from mean
weight between two
ultrasound
examinations
Marsoosi (2012)52 Prosp 43 EFW or AC < 10th Range, 23–40; MCA-PI/UA-PI Range, 1–30 days Perinatal death, IVH
(Iran) percentile for GA median, 31 < 1.08 and
and UA-PI and MCA-RI/
UA-RI > 2 SD above UA-RI < 1.0
mean for GA
Odibo (2014)53 Prosp 66 EFW < 10th percentile Range, 26–36; MCA-PI/UA-PI Adverse perinatal outcome (any of:
CPR predicts perinatal death in suspected FGR

≤ 1 week
(USA) for GA and mean, 31.1 < 1.08 perinatal death, 5-min Apgar
abnormal UA score < 3, cord arterial pH < 7.2,
Doppler (PI >95th seizures, NEC, grade 3 or 4 IVH,

Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.

percentile for GA or PVL)
absent/reversed
end-diastolic flow)
Lobmaier (2014)54 Prosp 198 EFW < 10th percentile Median, 34.0; MCA-PI/UA-PI 2–3 weeks Adverse perinatal outcome (operative
(Spain) for GA IQR, < 5th percentile delivery for NRFS or neonatal
32.9–36.3 metabolic acidosis (UA pH < 7.15
and base excess > 12 mEq/L))
Spinillo (2014)55 Prosp 176 AC < 10th percentile Mean, 32.6 and MCA-PI/UA-PI Mean, 21.5 and Perinatal death, mechanical ventilation,
(Italy) for GA on at least 30.6 in ≤10th percentile 17.7 days in normal neonatal brain lesion (severe IVH or
two consecutive normal and and abnormal CPR cystic PVL), CD
measurements, abnormal groups, respectively
2 weeks apart CPR groups,
respectively
Flood (2014)56 Prosp 881 EFW < 10th percentile Range, 24–36; MCA-PI/UA-PI Mean, 7 days; IQR, Adverse perinatal outcome (any of:
(Ireland) for GA mean, 30.1 and MCA-RI/ 2–15 days IVH, PVL, hypoxic ischemic
UA-RI < 1.0, encephalopathy, NEC, BPD, sepsis,
< 1.08, and death), admission to NICU, perinatal
< 5th percentile death
for GA
Crimmins (2015)57 Retro 192 EFW < 10th percentile > 32 MCA-PI/UA-PI ≤ 7 days Adverse perinatal outcome (arterial
(USA) for GA > 2 SD below pH < 7.0, base excess > −12 mEq/L
mean for GA or fetal death)
Babic (2015)58 Prosp 88 EFW < 10th percentile < 34 and ≥ 34 MCA-PI/UA-PI Median, 11.5 and Adverse perinatal outcome (any of: CD
(Canada) for GA < 5th percentile 14 days in normal for NRFS, 5-min Apgar score < 7,
and abnormal CPR NICU admission, perinatal death) and
groups, respectively its individual components, NICU stay

Ultrasound Obstet Gynecol 2018; 52: 430–441.

435

> 24 h

Continued over
 Table 1 Continued
436

GA at diagnosis
Definition of of suspected Abnormal test Interval between
Study (country) Design n suspected FGR FGR (weeks) result definition CPR and delivery Reference standard outcome
59 th
Figueras (2015) Prosp 509 EFW < 10 percentile > 32 MCA-PI/UA-PI ≤ 7 days Adverse perinatal outcome (CD for
(Spain) for GA < 10th percentile NRFS or neonatal metabolic acidosis
(UA pH < 7.15 and base excess
> −12 mEq/L)) and its individual
components, admission to NICU,
5-min Apgar score < 7, SGA at birth
Regan (2015)60 Retro 270 EFW < 10th percentile Mean, 31.2 and MCA-PI/UA-PI Mean, 5.7 and Adverse perinatal outcome (any of:
(USA) and/or AC < 5th 27.1 in ≤ 1.08 3.6 weeks in normal NICU admission, 5-min Apgar
percentile for GA normal and and abnormal CPR score < 5, UA pH < 7.10, meconium
abnormal groups, respectively passage, perinatal death), NICU
CPR groups, admission, perinatal death, CD for
respectively NRFS
Warshak (2015)61 Retro 154 EFW < 10th percentile < 32 MCA-PI/UA-PI Mean, 9.0 and Adverse perinatal outcome (any of:
(USA) and/or AC < 5th ≤ 1.08 4.0 weeks in normal NICU admission, 5-min Apgar
percentile for GA and abnormal CPR score < 5, UA pH < 7.10, meconium
groups, respectively passage, perinatal death), CD for
NRFS

Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.

Garcia-Simon (2015)62 Prosp 164 EFW < 10th percentile > 37 MCA-PI/UA-PI ∼48 h Adverse perinatal outcome (CD for fetal
(Spain) for GA < 5th percentile distress, neonatal metabolic acidosis
for GA (UA pH < 7.15 and base excess
> −12 mEq/L)) and its individual
components, CD, admission to NICU
Starčević (2016)63 Prosp 60 EFW < 10th percentile > 34 MCA-RI/UA-RI Unreported Neonatal brain lesion (PVL or peri/IVH)
(Croatia) for GA and increased ≤ 1.0 and ≤ 1.13 within 7 days after birth, and
UA-RI for GA abnormal functional neurological
status (as assessed by ATNAT
instrument) within 48 h of birth
Sirico (2018)64 Retro 310 EFW < 10th percentile Median, 37.3; Unreported Median, 4.2 days; IQR, UA pH < 7.1, 5-min Apgar score < 7,
(Germany) for GA IQR, 2.1–14.7 days pathological cardiotocograph trace,
35.4–39.0 presence of meconium-stained
amniotic fluid

Only first author of each study is given. *Median gestational age (GA) at delivery, 30 (range, 24–39) weeks. †GA at delivery, 31–40 weeks. ‡USA, Germany and UK. §Median GA at delivery, 31.2
(range, 24–36) weeks. AC, abdominal circumference; ATNAT, Amiel–Tison Neurological Assessment at Term; BPD, bronchopulmonary dysplasia; CD, Cesarean delivery; EFW, estimated fetal
weight; IQR, interquartile range; IVH, intraventricular hemorrhage; MCA, middle cerebral artery; NEC, necrotizing enterocolitis; NICU, neonatal intensive care unit; NRFS, non-reassuring fetal
status; PI, pulsatility index; Prosp, prospective; PVL, periventricular leukomalacia; RDS, respiratory distress syndrome; Retro, retrospective; RI, resistance index; S/D, systolic/diastolic; UA, umbilical
artery.
Conde-Agudelo et al.

Ultrasound Obstet Gynecol 2018; 52: 430–441.

CPR predicts perinatal death in suspected FGR 437

(20 011) 50
5) 62
) 3
) 9
) 4

) 7
92 4
4) 54
08 4
) 2

00 4

01
) 3

15 5
Sta ( 2 0 0 ) 1

12 5

) 9
16 6

) 8
5 6

2
4) 55

15 5
07 4

z(
) 53
) 7
) 60
1) 45

) 6

r as ) 6
) 4

) 8
ar s 0 0 5 4

(
01
03 4

(19
(20

14 5
18 6

(20

15 5
r ne (201

on

ne
20
(20

4
1

(20
r (2

) 1
20
Od (201
(20

Od (201

r tí
11 5
im
si (

ini
ur a

ed
20

ns
(20
(20

(20
ić (
vić

(2

a
ak

aie

a- S
ell
hs e

-M
oo

mi
llo

h(
og

(20
rče

ibo
ibo
rsh

ov
bm

od
am
ico

r ci

bic
ue

im
ini

ga

uz
an
ak
ak

Jug
Wa

Flo
Fig
Ste

Ga
Sir
Re

Lo

Gr

Fu

Cr
Cr

Ba
Sp

M
M
M
Study design
Description of the test
Selection of test cut-off value
Blinding of clinicians to test result
Inclusion of participants in the analysis
Use of interventions based on test results

Figure 2 Risk of bias of included studies reporting on prediction of adverse perinatal outcome by cerebroplacental ratio in singleton
pregnancy with fetal growth restriction suspected antenatally, according to Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2)
tool. Only first author of each study is given. , low risk of bias; , high risk of bias; , unclear risk of bias.

1.0 estimated pretest probabilities and summary positive and

negative LRs, we calculated that an abnormal CPR result
0.9 would increase the pretest probability of the composite
of adverse perinatal outcomes, perinatal death and SGA
0.8
at birth from 25% to 45%, 2% to 7.4% and 90% to
0.7 98.5%, respectively, whereas a normal CPR result would
decrease the pretest probability to 17%, 0.2% and 84%,
0.6 respectively.
Sensitivity

Visual assessment of both forest plots (Figure S1) and

0.5
summary ROC curves (Figure 3) suggested substantial
0.4
between-study heterogeneity, mainly for perinatal death,
any composite of adverse perinatal outcomes, Cesarean
0.3 delivery for non-reassuring fetal status and admission to
the NICU. Meta-regression analyses showed that none of
0.2 the prespecified covariates explained the heterogeneity
(Table S1). Sensitivity analyses revealed that pooled
0.1
predictive accuracy estimates obtained from studies with
0
low risk of bias in ≥ 5 domains did not differ significantly
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 from those obtained in the overall analysis (data not
Specificity shown). The funnel plot of the meta-analysis that included
at least 10 studies showed no significant asymmetry
Figure 3 Summary receiver–operating characteristics curves of (Deeks’ test P = 0.19).
cerebroplacental ratio for predicting composite of adverse perinatal
outcomes ( , ), perinatal death ( , ), Cesarean delivery
for non-reassuring fetal status ( , ), admission to neonatal
Subgroup analyses
intensive care unit ( , ), 5-min Apgar score < 7 ( , ) and
neonatal acidosis ( , ) in singleton pregnancies with fetal
Subgroup analyses of the accuracy of CPR to predict any
growth restriction suspected antenatally.
composite of adverse perinatal outcomes are depicted
in Table 3. CPR had a significantly higher predictive
ventilation, with summary positive and negative LRs that accuracy for any composite of adverse perinatal outcomes
varied between 1.1 and 2.5, and 0.3 and 0.9, respectively. among pregnancies with suspected early-onset FGR than
An abnormal CPR result had moderate accuracy for among those with suspected late-onset FGR. Moreover,
predicting SGA at birth (summary positive LR of 7.4; the accuracy of CPR for predicting any composite
two studies including 554 fetuses). Based on all included of adverse perinatal outcomes was lower when using
studies, we estimated that fetuses with suspected growth MCA-PI/UA-PI < 5th percentile as the definition of an
restriction had a prevalence rate (pretest probability) of abnormal result compared with other definitions, and
25% for the composite of adverse perinatal outcomes, when the CPR results were used to manage pregnancies
2.0% for perinatal death and 90% for SGA at birth compared with when they were not. There were no dif-
(birth weight < 10th percentile for GA). Then, based on ferences in the predictive ability of CPR between studies

Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd. Ultrasound Obstet Gynecol 2018; 52: 430–441.
438 Conde-Agudelo et al.

Table 2 Accuracy of cerebroplacental ratio for predicting adverse perinatal outcome in singleton pregnancy with fetal growth restriction
suspected antenatally

Pooled Pooled
sensitivity specificity LR+ LR−
Outcome Studies (n) Women (n) (% (95% CI)) (% (95% CI)) (95% CI) (95% CI)

Perinatal death 645,52,55,56,58,60 1495 93 (78–98) 76 (74–78) 3.9 (3.4–4.5) 0.09 (0.0–0.3)
Any composite of 1143,44,47,53,54,56–60,62 2658 57 (53–61) 77 (75–79) 2.5 (2.3–2.8) 0.6 (0.5–0.6)
adverse perinatal
outcomes
CD for NRFS 743,45,50,58–60,62 1339 59 (54–64) 74 (72–77) 2.3 (2.0–2.6) 0.6 (0.5–0.6)
5-min Apgar score < 7 643,44,51,58,59,64 1148 54 (42–66) 72 (69–74) 1.9 (1.5–2.4) 0.6 (0.5–0.8)
Admission to NICU 943–45,51,56,58–60,62 2206 45 (41–49) 79 (77–81) 2.2 (1.9–2.5) 0.7 (0.6–0.8)
Neonatal acidosis 550,51,59,62,64 1283 48 (38–58) 70 (68–73) 1.6 (1.3–2.0) 0.7 (0.6–0.9)
Neonatal brain lesion 545,48,52,55,63 352 56 (43–67) 48 (43–54) 1.1 (0.8–1.4) 0.9 (0.7–1.2)
Neonatal morbidity 443,45,46,49 547 78 (67–86) 33 (29–37) 1.2 (1.0–1.3) 0.7 (0.4–1.1)
other than brain
lesion
Use of mechanical 155 176 90 (77–96) 39 (31–47) 1.5 (1.2–1.7) 0.3 (0.1–0.7)
ventilation
SGA at birth* 243,59 554 43 (39–47) 94 (84–98) 7.4 (2.5–22.4) 0.6 (0.5–0.7)

*Birth weight < 10th percentile. CD, Cesarean delivery; NICU, neonatal intensive care unit; NRFS, non-reassuring fetal status; LR−,
negative likelihood ratio; LR+, positive likelihood ratio; SGA, small-for-gestational age.

Table 3 Subgroup analyses of predictive accuracy of cerebroplacental ratio (CPR) for any composite of adverse perinatal outcome in
singleton pregnancy with fetal growth restriction (FGR) suspected antenatally

Pooled Pooled
Women sensitivity specificity LR+ LR−
Subgroup Studies (n) (n) (% (95% CI)) (% (95% CI)) (95% CI) (95% CI)

All studies 1143,44,47,53,54,56–60,62 2658 57 (53–61) 77 (75–79) 2.5 (2.3–2.8) 0.6 (0.5–0.6)
Onset of FGR
Early (< 32 or < 34 weeks 347,56,61 784 64 (56–71) 85 (82–88) 4.2 (3.4–5.3) 0.4 (0.3–0.5)
at diagnosis or delivery)
Late (≥ 32 or ≥ 34 weeks at 647,54,56,57,59,62 1915 56 (51–61) 76 (74–78) 2.3 (2.0–2.6) 0.6 (0.5–0.7)
diagnosis or delivery)
Definition of abnormal CPR
MCA-PI/UA-PI ≤ 1.08 543,47,53,56,60 1445 59 (52–65) 81 (79–83) 3.1 (2.7–3.6) 0.5 (0.4–0.6)
MCA-PI/UA-PI < 5th percentile 454,56,58,62 1323 63 (56–69) 62 (59–65) 1.7 (1.5–1.9) 0.6 (0.5–0.7)
MCA-RI/UA-RI < 1 or < 1.05 244,56 942 63 (53–73) 84 (81–86) 3.9 (3.1–4.8) 0.4 (0.3–0.6)
Interval from CPR to delivery
≤ 7 days 553,56,57,59,62 1812 62 (56–67) 76 (73–78) 2.5 (2.2–2.9) 0.5 (0.4–0.6)
> 7 days 544,47,54,58,60 801 53 (47–58) 82 (78–85) 2.9 (2.4–3.6) 0.6 (0.5–0.7)
Definition of suspected FGR
based on:
Only EFW < 10th percentile 844,47,54,56–59,62 2277 60 (56–65) 75 (73–77) 2.4 (2.2–2.7) 0.5 (0.5–0.6)
for GA
EFW < 10th percentile for GA 153 66 59 (36–78) 76 (62–85) 2.4 (1.3–4.5) 0.5 (0.3–1.0)
and abnormal UA Doppler
Management of pregnancy
Not using CPR results 943,44,47,53,56,58–60,62 2268 58 (54–62) 80 (78–82) 2.9 (2.5–3.2) 0.5 (0.5–0.6)
Using CPR results 154 198 49 (38–61) 74 (66–81) 1.9 (1.3–2.8) 0.7 (0.5–0.9)
Low risk of bias in 643,53,56,58,59,62 1745 61 (56–66) 78 (76–80) 2.8 (2.5–3.2) 0.5 (0.4–0.6)
≥ 5 domains

EFW, estimated fetal weight; GA, gestational age; LR−, negative likelihood ratio; LR+, positive likelihood ratio; MCA, middle cerebral
artery; PI, pulsatility index; RI, resistance index; UA, umbilical artery.

in which the interval from CPR measurement to delivery DISCUSSION

was ≤ 7 days and those in which it was > 7 days, and
between studies using EFW < 10th percentile for GA as the
definition of suspected FGR and those using EFW < 10th
percentile for GA and abnormal UA Doppler as the
definition.
Main findings
The results of this systematic review indicate that CPR
has moderate-to-high predictive accuracy for perinatal
death, the most important outcome measure in relation to

Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd. Ultrasound Obstet Gynecol 2018; 52: 430–441.
CPR predicts perinatal death in suspected FGR
uteroplacental insufficiency in suspected FGR. In partic-
ular, a normal CPR result had high accuracy to identify
which fetuses with suspected growth restriction are at low
risk of dying in the perinatal period, decreasing the pretest
probability of perinatal death from 2% to 0.2%. Overall,
CPR had low predictive accuracy for the other adverse
perinatal outcomes considered, several of which are less
well correlated with uteroplacental insufficiency in sus-
pected FGR. Notwithstanding, the presence of an abnor-
mal CPR result increased the pretest probability of adverse
perinatal outcome from 25% to 45%. In addition, sub-
group analyses suggest that the predictive accuracy of CPR
is higher in pregnancies with suspected early-onset FGR
and when MCA-PI/UA-PI ≤ 1.08 or MCA-RI/UA-RI < 1
or < 1.05 is used as the definition of abnormal CPR.
Previously, it has been suggested that CPR is a stronger
predictor of adverse perinatal outcome in suspected
late-onset FGR than in suspected early-onset FGR23,25–27 .
Unexpectedly, our subgroup analysis showed the oppo-
site; there was a higher predictive accuracy for
adverse perinatal outcome in pregnancies with suspected
early-onset FGR. Usually, suspected late-onset FGR is
characterized by abnormal Doppler indices involving the
MCA, with normal or minimally elevated resistance in
the UA22 . In contrast, suspected early-onset FGR is char-
acterized by abnormal Doppler indices of both the UA
and MCA22 . Abnormality in both vessels being included
in the calculation of CPR, in particular high values of
UA Doppler indices, could explain the better predictive
accuracy of CPR in suspected early-onset FGR in com-
parison with suspected late-onset FGR in which there are
abnormal indices in only one vessel.
It is noteworthy that no included study provided data
to assess the predictive ability of CPR for adverse neuro-
developmental outcome in pregnancies with suspected
FGR. A secondary analysis of the TRUFFLE study65
reported that CPR was not associated with neurodevelop-
mental impairment at 2 years’ corrected age in fetuses with
suspected early-onset growth restriction66 . Two studies
reported similar results for decreased MCA-PI in sus-
pected FGR67,68 . A systematic review reported that SGA
or growth-restricted fetuses with cerebral redistribution
may be at higher risk of adverse neurodevelopmental
outcome69 . However, none of the studies included in this
review used CPR for defining cerebral redistribution.
Strengths and limitations
The reliability and robustness of our systematic review
are supported by: (1) adherence to guidelines for
the conduct and reporting of systematic reviews of
predictive test accuracy; (2) use of a prospective protocol
designed to address a highly specific research question;
(3) comprehensive literature search without language
restrictions; (4) inclusion of a relatively large number
of studies, mostly published in recent years; (5) strict
study quality assessment; (6) quantitative synthesis of the
evidence; (7) use of contemporary statistical methods
to obtain summary measures of predictive accuracy
Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.
439
including subgroup and sensitivity analyses; and (8)
exploration of potential sources of heterogeneity.
Limitations include lack of blinding to CPR results or
omission of information on this subject in approximately
two-thirds of the included studies. Although most studies
reported that the CPR results were not used to manage
pregnancies with suspected FGR, it is possible that
women with an abnormal CPR result were followed
up more closely or received interventions, which could
have introduced bias in the assessment of the test’s
predictive accuracy. However, sensitivity analyses that
were restricted to studies at low risk of blinding bias
showed no significant differences in the results obtained
in overall meta-analyses.
There were considerable differences among stud-
ies in the definition of suspected FGR and Doppler
indices/cut-off values used for defining abnormal CPR,
which limit the generalizability of our findings. More-
over, prespecified variables did not explain the substantial
heterogeneity and, therefore, pooled estimates of predic-
tive accuracy should be interpreted cautiously. Finally, the
statistical power of some of our meta-analyses was limited
by the small number of studies within each subgroup and
the relatively small number of outcome events in some
included studies.
Interpretation in light of previous systematic reviews
We identified three systematic reviews on the predictive
accuracy of CPR for adverse perinatal outcome70–72 .
Nassr et al.70 included seven studies, and reported
that abnormal CPR in pregnancies at high risk for
FGR or with a diagnosis of FGR increased the risk
for adverse perinatal outcome. Summary ROC curves
showed that CPR had a better predictive accuracy
for neonatal complications and NICU admission. Dunn
et al.71 reported that CPR was predictive of Cesarean
section for intrapartum fetal compromise, SGA and
NICU admission in pregnancies at term. These reviews
did not report pooled estimates of predictive accuracy.
Finally, Vollgraff Heidweiller-Schreurs et al.72 assessed
the accuracy of CPR to predict adverse perinatal outcome
in singleton pregnancies of all risk profiles. CPR was
significantly superior to UA and MCA Doppler in
predicting a composite of adverse perinatal outcomes
and emergency delivery for fetal distress. No differences
were found between CPR and either UA Doppler or MCA
Doppler in the prediction of perinatal death, low Apgar
score or NICU admission. Overall, our estimates of the
predictive accuracy of CPR for adverse perinatal outcome
among pregnancies with suspected FGR were lower than
those reported in this review among pregnancies of all
risk profiles.
CPR has been hypothesized to be a more accurate
test for predicting adverse perinatal outcome than its
individual components, UA and MCA Doppler. When
comparing the estimates obtained in our study with
those reported in two meta-analyses that assessed the
accuracy of UA73 and MCA74 Doppler to predict adverse
Ultrasound Obstet Gynecol 2018; 52: 430–441.
440
perinatal outcome in high-risk pregnancies, CPR had
better predictive accuracy for perinatal death (summary
positive and negative LRs of 3.9 and 0.09, respectively)
than both UA Doppler (summary positive and negative
LRs of 2.5 and 0.3, respectively) and MCA Doppler
(summary positive and negative LRs of 1.4 and 0.5,
respectively). In general, the predictive accuracy of CPR
for other adverse perinatal outcomes appeared to be
comparable to those of UA and MCA Doppler.
Conclusions
CPR appears to be useful in predicting perinatal death
in pregnancies with suspected FGR. Nevertheless, before
incorporating CPR into the routine clinical management
of suspected FGR, randomized controlled trials, ideally
blinded, should assess whether the use of CPR reduces
perinatal death or other adverse perinatal outcomes.
Studies are required to assess the predictive accuracy of
CPR for adverse neurodevelopmental outcome in fetuses
with growth restriction suspected antenatally.
ACKNOWLEDGMENTS
A.T.P. is supported by the National Institute for Health
Research (NIHR) Oxford Biomedical Research Centre
(BRC). We are very grateful to Drs Francesc Figueras, Inas
Babic and Carri R. Warshak for providing unpublished
data from their studies and for clarifying other queries.
This work was supported by departmental funds.
REFERENCES
1. de Onis M, Blössner M, Villar J. Levels and patterns of intrauterine growth
retardation in developing countries. Eur J Clin Nutr 1998; 52 (Suppl 1): S5–15.
2. Lee AC, Katz J, Blencowe H, Cousens S, Kozuki N, Vogel JP, Adair L, Baqui AH,
Bhutta ZA, Caulfield LE, Christian P, Clarke SE, Ezzati M, Fawzi W, Gonzalez
R, Huybregts L, Kariuki S, Kolsteren P, Lusingu J, Marchant T, Merialdi M,
Mongkolchati A, Mullany LC, Ndirangu J, Newell ML, Nien JK, Osrin D, Roberfroid
D, Rosen HE, Sania A, Silveira MF, Tielsch J, Vaidya A, Willey BA, Lawn JE, Black
RE; CHERG SGA-Preterm Birth Working Group. National and regional estimates
of term and preterm babies born small for gestational age in 138 low-income and
middle-income countries in 2010. Lancet Glob Health 2013; 1: e26–e36.
3. WHO Expert Committee on Physical Status. Physical Status: The use and
Interpretation of Anthropometry – Report of a WHO Expert Committee. WHO
Technical Report Series 854. Geneva: World Health Organization, 1995.
4. Katz J, Lee AC, Kozuki N, Lawn JE, Cousens S, Blencowe H, Ezzati M, Bhutta
ZA, Marchant T, Willey BA, Adair L, Barros F, Baqui AH, Christian P, Fawzi
W, Gonzalez R, Humphrey J, Huybregts L, Kolsteren P, Mongkolchati A, Mullany
LC, Ndyomugyenyi R, Nien JK, Osrin D, Roberfroid D, Sania A, Schmiegelow C,
Silveira MF, Tielsch J, Vaidya A, Velaphi SC, Victora CG, Watson-Jones D, Black RE;
CHERG Small-for-Gestational-Age-Preterm Birth Working Group. Mortality risk in
preterm and small-for-gestational-age infants in low-income and middle-income
countries: a pooled country analysis. Lancet 2013; 382: 417–425.
5. Hirst JE, Villar J, Victora CG, Papageorghiou AT, Finkton D, Barros FC, Gravett
MG, Giuliani F, Purwar M, Frederick IO, Pang R, Cheikh Ismail L, Lambert A,
Stones W, Jaffer YA, Altman DG, Noble JA, Ohuma EO, Kennedy SH, Bhutta
ZA; International Fetal and Newborn Growth Consortium for the 21st Century
(INTERGROWTH-21st). The antepartum stillbirth syndrome: risk factors and
pregnancy conditions identified from the INTERGROWTH-21st Project. BJOG
2018; 125: 1145–1153.
6. Longo S, Borghesi A, Tzialla C, Stronati M. IUGR and infections. Early Hum Dev
2014; 90 (Suppl 1): S42–44.
7. Chernausek SD. Update: consequences of abnormal fetal growth. J Clin Endocrinol
Metab 2012; 97: 689–695.
8. Baschat AA. Neurodevelopment after fetal growth restriction. Fetal Diagn Ther
2014; 36: 136–142.
9. Murray E, Fernandes M, Fazel M, Kennedy SH, Villar J, Stein A. Differential effect of
intrauterine growth restriction on childhood neurodevelopment: a systematic review.
BJOG 2015; 122: 1062–1072.
Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.
Conde-Agudelo et al.
10. Barker DJ. Adult consequences of fetal growth restriction. Clin Obstet Gynecol
2006; 49: 270–283.
11. Salam RA, Das JK, Bhutta ZA. Impact of intrauterine growth restriction on long-term
health. Curr Opin Clin Nutr Metab Care 2014; 17: 249–254.
12. American College of Obstetricians and Gynecologists. ACOG Practice Bulletin no.
134: Fetal growth restriction. Obstet Gynecol 2013; 121: 1122–1133.
13. Royal College of Obstetricians and Gynaecologists (RCOG). The Investigation and
Management of the Small-for-Gestational-Age Fetus. Green-top Guideline No. 31.
London: RCOG, 2013.
14. Society for Maternal-Fetal Medicine Publications Committee, Berkley E, Chauhan SP,
Abuhamad A. Doppler assessment of the fetus with intrauterine growth restriction.
Am J Obstet Gynecol 2012; 206: 300–308.
15. Lausman A, Kingdom J; Maternal Fetal Medicine Committee. Intrauterine growth
restriction: screening, diagnosis, and management. J Obstet Gynaecol Can 2013; 35:
741–748.
16. Gordijn SJ, Beune IM, Thilaganathan B, Papageorghiou A, Baschat AA, Baker PN,
Silver RM, Wynia K, Ganzevoort W. Consensus definition of fetal growth restriction:
a Delphi procedure. Ultrasound Obstet Gynecol 2016; 48: 333–339.
17. Chauhan SP, Dahlke JD, Magann EF, Chang E, Gupta L, Mol BW, Lewis DF. Isolated
intrauterine growth restriction: a survey of Central Association of Obstetricians
Gynecologists (CAOG) members. J Matern Fetal Neonatal Med 2013; 26: 497–502.
18. Unterscheider J, Daly S, Geary MP, Kennelly MM, McAuliffe FM, O’Donoghue K,
Hunter A, Morrison JJ, Burke G, Dicker P, Tully EC, Malone FD. Definition and
management of fetal growth restriction: a survey of contemporary attitudes. Eur J
Obstet Gynecol Reprod Biol 2014; 174: 41–45.
19. Savchev S, Figueras F, Gratacos E. Survey on the current trends in managing
intrauterine growth restriction. Fetal Diagn Ther 2014; 36: 129–135.
20. Alfirevic Z, Stampalija T, Dowswell T. Fetal and umbilical Doppler ultrasound in
high-risk pregnancies. Cochrane Database Syst Rev 2017; 6: CD007529.
21. Arbeille P, Roncin A, Berson M, Patat F, Pourcelot L. Exploration of the fetal cerebral
blood flow by duplex Doppler—linear array system in normal and pathological
pregnancies. Ultrasound Med Biol 1987; 13: 329–337.
22. DeVore GR. The importance of the cerebroplacental ratio in the evaluation of fetal
well-being in SGA and AGA fetuses. Am J Obstet Gynecol 2015; 213: 5–15.
23. Hernandez-Andrade E, Benavides Serralde JA, Cruz-Martinez R. Can anomalies of
fetal brain circulation be useful in the management of growth restricted fetuses?
Prenat Diagn 2012; 32: 103–112.
24. Seravalli V, Baschat AA. A uniform management approach to optimize outcome in
fetal growth restriction. Obstet Gynecol Clin North Am 2015; 42: 275–288.
25. Morales-Roselló J, Khalil A. Fetal cerebral redistribution: a marker of compromise
regardless of fetal size. Ultrasound Obstet Gynecol 2015; 46: 385–388.
26. Khalil A, Thilaganathan B. Role of uteroplacental and fetal Doppler in identifying
fetal growth restriction at term. Best Pract Res Clin Obstet Gynaecol 2017; 38:
38–47.
27. Figueras F, Gratacos E. An integrated approach to fetal growth restriction. Best Pract
Res Clin Obstet Gynaecol 2017; 38: 48–58.
28. Nawathe A, Lees C. Early onset fetal growth restriction. Best Pract Res Clin Obstet
Gynaecol 2017; 38: 24–37.
29. Dall’Asta A, Brunelli V, Prefumo F, Frusca T, Lees CC. Early onset fetal growth
restriction. Matern Health Neonatol Perinatol 2017; 3: 2.
30. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, Lijmer JG,
Moher D, Rennie D, de Vet HC; Standards for Reporting of Diagnostic Accuracy.
Towards complete and accurate reporting of studies of diagnostic accuracy: the
STARD initiative. BMJ 2003; 326: 41–44.
31. Leeflang MM, Deeks JJ, Gatsonis C, Bossuyt PM; Cochrane Diagnostic Test Accuracy
Working Group. Systematic reviews of diagnostic test accuracy. Ann Intern Med
2008; 149: 889–897.
32. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, Leeflang
MM, Sterne JA, Bossuyt PM; QUADAS-2 Group. QUADAS-2: a revised tool for
the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011; 155:
529–536.
33. Whiting P, Harbord R, Kleijnen J. No role for quality scores in systematic reviews of
diagnostic accuracy studies. BMC Med Res Methodol 2005; 5: 19.
34. Sankey S, Weisfiels L, Fine M, Kapoor W. An assessment of the use of the continuity
correction for sparse data in meta-analysis. Commun Stat Simulation Computation
1996; 25: 1031–1056.
35. Rutter CM, Gatsonis CA. A hierarchical regression approach to meta-analysis of
diagnostic test accuracy evaluations. Stat Med 2001; 20: 2865–2884.
36. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH.
Bivariate analysis of sensitivity and specificity produces informative summary
measures in diagnostic reviews. J Clin Epidemiol 2005; 58: 982–990.
37. Walter SD. Properties of the summary receiver operating characteristic (SROC) curve
for diagnostic test data. Stat Med 2002; 21: 1237–1256.
38. Zwinderman AH, Bossuyt PM. We should not pool diagnostic likelihood ratios in
systematic reviews. Stat Med 2008; 27: 687–697.
39. Jaeschke R, Guyatt GH, Sackett DL; The Evidence-Based Medicine Working Group.
Users’ guides to the medical literature. III. How to use an article about a diagnostic
test. B. What are the results and will they help me in caring for my patients? JAMA
1994; 271: 703–707.
40. Macaskill P, Gatsonis C, Deeks JJ, Harbord RM, Takwoingi Y. Analysing and
presenting results. In Cochrane Handbook for Systematic Reviews of Diagnostic
Test Accuracy Version 1.0, Deeks JJ, Bossuyt PM, Gatsonis C (eds), Chapter 10. The
Cochrane Collaboration: London, 2010.
41. Lijmer JG, Bossuyt PM, Heisterkamp SH. Exploring sources of heterogeneity in
systematic reviews of diagnostic tests. Stat Med 2002; 21: 1525–1537.
42. Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other
sample size effects in systematic reviews of diagnostic test accuracy was assessed. J
Clin Epidemiol 2005; 58: 882–893.
Ultrasound Obstet Gynecol 2018; 52: 430–441.
CPR predicts perinatal death in suspected FGR
43. Gramellini D, Folli MC, Raboni S, Vadora E, Merialdi A. Cerebral-umbilical Doppler
ratio as a predictor of adverse perinatal outcome. Obstet Gynecol 1992; 79: 416–420.
44. Makhseed M, Jirous J, Ahmed MA, Viswanathan DL. Middle cerebral artery to
umbilical artery resistance index ratio in the prediction of neonatal outcome. Int J
Gynaecol Obstet 2000; 71: 119–125.
45. Sterne G, Shields LE, Dubinsky TJ. Abnormal fetal cerebral and umbilical Doppler
measurements in fetuses with intrauterine growth restriction predicts the severity of
perinatal morbidity. J Clin Ultrasound 2001; 29: 146–151.
46. Makh DS, Harman CR, Baschat AA. Is Doppler prediction of anemia effective in the
growth-restricted fetus? Ultrasound Obstet Gynecol 2003; 22: 489–492.
47. Odibo AO, Riddick C, Pare E, Stamilio DM, Macones GA. Cerebroplacental Doppler
ratio and adverse perinatal outcomes in intrauterine growth restriction: evaluating
the impact of using gestational age-specific reference values. J Ultrasound Med 2005;
24: 1223–1228.
48. Jugović D, Tumbri J, Medić M, Jukić MK, Kurjak A, Arbeille P, Salihagić-Kadić A.
New Doppler index for prediction of perinatal brain damage in growth-restricted
and hypoxic fetuses. Ultrasound Obstet Gynecol 2007; 30: 303–311.
49. Manogura AC, Turan O, Kush ML, Berg C, Bhide A, Turan S, Moyano D, Bower
S, Nicolaides KH, Galan HL, Müller T, Thilaganathan B, Gembruch U, Harman
CR, Baschat AA. Predictors of necrotizing enterocolitis in preterm growth-restricted
neonates. Am J Obstet Gynecol 2008; 198: 638.e1–638.e5.
50. Cruz-Martı́nez R, Figueras F, Hernandez-Andrade E, Oros D, Gratacos E. Fetal
brain Doppler to predict cesarean delivery for nonreassuring fetal status in term
small-for-gestational-age fetuses. Obstet Gynecol 2011; 117: 618–626.
51. Fu J, Olofsson P. Relations between fetal brain-sparing circulation, oxytocin challenge
test, mode of delivery and fetal outcome in growth-restricted term fetuses. Acta Obstet
Gynecol Scand 2011; 90: 227–230.
52. Marsoosi V, Bahadori F, Esfahani F, Ghasemi-Rad M. The role of Doppler indices
in predicting intra ventricular hemorrhage and perinatal mortality in fetal growth
restriction. Med Ultrason 2012; 14: 125–132.
53. Odibo AO, Goetzinger KR, Cahill AG, Odibo L, Macones GA. Combined
sonographic testing index and prediction of adverse outcome in preterm fetal growth
restriction. Am J Perinatol 2014; 31: 139–144.
54. Lobmaier SM, Figueras F, Mercade I, Perello M, Peguero A, Crovetto F, Ortiz JU,
Crispi F, Gratacós E. Angiogenic factors vs Doppler surveillance in the prediction of
adverse outcome among late-pregnancy small-for-gestational-age fetuses. Ultrasound
Obstet Gynecol 2014; 43: 533–540.
55. Spinillo A, Gardella B, Bariselli S, Alfei A, Silini EM, Bello BD. Cerebroplacental
Doppler ratio and placental histopathological features in pregnancies complicated by
fetal growth restriction. J Perinat Med 2014; 42: 321–328.
56. Flood K, Unterscheider J, Daly S, Geary MP, Kennelly MM, McAuliffe FM,
O’Donoghue K, Hunter A, Morrison JJ, Burke G, Dicker P, Tully EC, Malone
FD. The role of brain sparing in the prediction of adverse outcomes in intrauterine
growth restriction: results of the multicenter PORTO Study. Am J Obstet Gynecol
2014; 211: 288.e1–288.e5.
57. Crimmins S, Fitzgerald G, Block-Abraham D, Atkins K, Harman C, Turan O. Use
of cerebroplacental ratio to predict outcome in late onset fetal growth restriction
(FGR). Am J Obstet Gynecol 2015; 212 (Suppl 1): S269–S270.
58. Babic I, Alameri S, Tulbah M, Moretti F, Kurdi W. Cerebroplacental ratio and
its association with adverse perinatal outcomes in early and late onset intrauterine
growth restriction (IUGR). Ultrasound Obstet Gynecol 2015; 46 (Suppl 1): 149.
59. Figueras F, Savchev S, Triunfo S, Crovetto F, Gratacos E. An integrated model with
classification criteria to predict small-for-gestational-age fetuses at risk of adverse
perinatal outcome. Ultrasound Obstet Gynecol 2015; 45: 279–285.
SUPPORTING INFORMATION ON THE INTERNET
The following supporting information may be found in the online version of this article:
Appendix S1 Search strategy for studies on predictive accuracy of cerebroplacental ratio for adverse perinatal
and neurodevelopmental outcomes in suspected fetal growth restriction
Appendix S2 Assessment of risk of bias in included studies
Table S1 Metaregression analyses to identify sources of between-study heterogeneity in meta-analyses of
predictive accuracy of cerebroplacental ratio (CPR) for adverse perinatal outcome in singleton pregnancy with
fetal growth restriction suspected antenatally
Figure S1 Forest plots of cerebroplacental ratio to predict adverse perinatal outcome in suspected fetal growth
restriction.
Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.
441
60. Regan J, Masters H, Warshak C. Association between an abnormal cerebroplacental
ratio and the development of severe pre-eclampsia. J Perinatol 2015; 35: 322–327.
61. Warshak CR, Masters H, Regan J, DeFranco E. Doppler for growth restriction: the
association between the cerebroplacental ratio and a reduced interval to delivery. J
Perinatol 2015; 35: 332–337.
62. Garcia-Simon R, Figueras F, Savchev S, Fabre E, Gratacos E, Oros D. Cervical
condition and fetal cerebral Doppler as determinants of adverse perinatal outcome
after labor induction for late-onset small-for-gestational-age fetuses. Ultrasound
Obstet Gynecol 2015; 46: 713–717.
63. Starčević M, Predojević M, Butorac D, Tumbri J, Konjevoda P, Kadić AS. Early
functional and morphological brain disturbances in late-onset intrauterine growth
restriction. Early Hum Dev 2016; 93: 33–38.
64. Sirico A, Diemert A, Glosemeyer P, Hecher K. Prediction of adverse perinatal
outcome by cerebroplacental ratio adjusted for estimated fetal weight. Ultrasound
Obstet Gynecol 2018; 51: 381–386.
65. Lees CC, Marlow N, van Wassenaer-Leemhuis A, Arabin B, Bilardo CM, Brezinka
C, Calvert S, Derks JB, Diemert A, Duvekot JJ, Ferrazzi E, Frusca T, Ganzevoort W,
Hecher K, Martinelli P, Ostermayer E, Papageorghiou AT, Schlembach D, Schneider
KT, Thilaganathan B, Todros T, Valcamonico A, Visser GH, Wolf H; TRUFFLE
study group. 2 year neurodevelopmental and intermediate perinatal outcomes in
infants with very preterm fetal growth restriction (TRUFFLE): a randomised trial.
Lancet 2015; 385: 2162–2172.
66. Stampalija T, Arabin B, Wolf H, Bilardo CM, Lees C; TRUFFLE investigators.
Is middle cerebral artery Doppler related to neonatal and 2-year infant
outcome in early fetal growth restriction? Am J Obstet Gynecol 2017; 216:
521.e1–521.e13.
67. 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.
68. Constantinescu S, Denes M, Vladareanu R. Intrauterine growth restriction: Perinatal
assessment in predicting the offspring neurologic impairment. A 2 years prospective
study. Ginecoeu 2013; 9: 35–40.
69. Meher S, Hernandez-Andrade E, Basheer SN, Lees C. Impact of cerebral
redistribution on neurodevelopmental outcome in small-for-gestational-age or
growth-restricted babies: a systematic review. Ultrasound Obstet Gynecol 2015;
46: 398–404.
70. Nassr AA, Abdelmagied AM, Shazly SA. Fetal cerebro-placental ratio and adverse
perinatal outcome: systematic review and meta-analysis of the association and
diagnostic performance. J Perinat Med 2016; 44: 249–256.
71. Dunn L, Sherrell H, Kumar S. Review: Systematic review of the utility of the fetal
cerebroplacental ratio measured at term for the prediction of adverse perinatal
outcome. Placenta 2017; 54: 68–75.
72. Vollgraff Heidweiller-Schreurs CA, de Boer MA, Heymans MW, Schoonmade
LJ, Bossuyt PMM, Mol BWJ, De Groot CJM, Bax CJ. Prognostic accuracy of
cerebroplacental ratio and middle cerebral artery Doppler for adverse perinatal
outcomes: a systematic review and meta-analysis. Ultrasound Obstet Gynecol 2018;
51: 313–322.
73. Morris RK, Malin G, Robson SC, Kleijnen J, Zamora J, Khan KS. Fetal umbilical
artery Doppler to predict compromise of fetal/neonatal wellbeing in a high-risk
population: systematic review and bivariate meta-analysis. Ultrasound Obstet
Gynecol 2011; 37: 135–142.
74. Morris RK, Say R, Robson SC, Kleijnen J, Khan KS. Systematic review and
meta-analysis of middle cerebral artery Doppler to predict perinatal wellbeing.
Eur J Obstet Gynecol Reprod Biol 2012; 165: 141–155.
Ultrasound Obstet Gynecol 2018; 52: 430–441.
Ultrasound Obstet Gynecol 2018; 52: 430–441
Published online 6 September 2018 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/uog.19117

Precisi ón en la predicci ón de la relaci ón cerebroplacentaria para los resultados adversos perina-
tales y del desarrollo neurol ógico en embarazos con sospecha de restricci ón del crecimiento fetal:
revisi ón sistem ática y metaan álisis
RESUMEN
Objetivo La relación cerebroplacentaria (RCP) se ha propuesto como el método para vigilar de forma sistemática
los embarazos con sospecha de restricción del crecimiento fetal (RCF), pero el rendimiento predictivo de esta prueba
es incierto. El objetivo de este estudio fue determinar la precisión de la RCP para predecir los resultados adversos
perinatales y del desarrollo neurológico en casos con sospecha de RCF.
Métodos Se realizaron búsquedas en PubMed, EMBASE, CINAHL y Lilacs desde su inicio hasta el 31 de julio de 2017,
en busca de estudios de cohortes o transversales que informasen sobre la precisión de la RCP para predecir resultados
adversos perinatales y/o del desarrollo neurológico en embarazos únicos con sospecha prenatal de RCF, a partir de
parámetros ecográficos. Se generaron curvas resumen de las caracterı́sticas operativas del receptor (ROC, por sus siglas
en inglés), de las sensibilidades y especificidades combinadas, y de los cocientes de verosimilitud (LRs).
Resultados Veintidós estudios (4 301 mujeres) cumplieron los criterios de inclusión. Las curvas resumen ROC mostraron
que la mejor precisión predictiva de la RCP fue para la muerte perinatal y la peor fue para la acidosis neonatal, con áreas
por debajo de las curvas resumen ROC de 0,83 y 0,57, respectivamente. La precisión predictiva de la RCP fue entre
moderada a alta para la muerte perinatal (sensibilidad y especificidad combinadas del 93% y el 76%, respectivamente, y
LRs positivos y negativos resumen del 3,9 y 0,09, respectivamente) y baja para la combinación de resultados perinatales
adversos, cesárea por riesgo de pérdida de bienestar fetal, puntuación de Apgar a los 5 minutos <7, ingreso a la unidad
de cuidados intensivos para neonatos, acidosis neonatal y morbilidad neonatal, con un resumen de LRs positivos y
negativos de 1,1 a 2,5 y de 0,3 a 0,9, respectivamente. Un resultado anómalo de RCP tuvo una precisión moderada
para predecir los nacimientos pequeños para la edad gestacional (resumen LR positivo 7,4). La RCP tuvo una mayor
precisión predictiva en los embarazos con sospecha de RCF de aparición temprana. Ningún estudio proporcionó datos
para evaluar la precisión predictiva de la RCP para resultados adversos del desarrollo neurológico.
Conclusión La RCP parece ser útil para predecir la muerte perinatal en embarazos con sospecha de RCF. Sin embargo,
antes de incorporar la RCP en el tratamiento clı́nico habitual de casos con sospecha de RCF, se deberı́a evaluar
mediante ensayos controlados aleatorizados si el uso de la RCP reduce la muerte perinatal u otros resultados perinatales
adversos.

Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd. SYSTEMATIC REVIEW