IMAGING
Quality control of an image-scoring method for nuchal
translucency ultrasonography
Nicolas Fries, MD; Marc Althuser, MD; Marianne Fontanges, MD; Claude Talmant, MD; Pierre S. Jouk, PhD;
Malka Tindel, SM; Michel Duyme, PhD
OBJECTIVE: The objective of the study was to describe a new image-
scoring method (NISM) for the measurement of nuchal translucency
and crown-rump length on ultrasound scans and to establish interre-
viewer reliability.
STUDY DESIGN: This NISM was based on 8 criteria on a scale of 4
levels (1-4) established on clearly defined ultrasound reference marks.
Ten reviewers assessed the same images of 30 fetuses. After a short
training period, the same images and those of 30 new fetuses were
scored by these 10 reviewers.
RESULTS: The differences in scores among the 10 reviewers were sig-
nificant for 4 of 45 pairwise comparisons before training, but no pair-
wise comparison was significant after training. Interreviewer variance
was significantly lower after training (P = .045). The intraclass corre-
lations before and after training were 0.75 and 0.82. For each criterion,
the scores were dichotomized into 2 categories (1-2 vs 3-4). Kappa
values for each criterion were substantial (0.61 to 0.80) or even almost
perfect (0.81 to 1.00).
CONCLUSION: This NISM was highly reliable for the total scores and
for each criterion evaluating the image of nuchal translucency and
crown-rump length and provides a relevant quality control tool for ul-
trasound operators.
Key words: Down’s syndrome, ongoing audit, 11 to 14 week scan,
ultrasound screening
Cite this article as: Fries N, Althuser M, Fontanges M, et al. Quality control of an image-scoring method for nuchal translucency ultrasonography. Am J Obstet
Gynecol 2007;196:272.e1-272.e5.
M
easurements of nuchal translu-
cency (NT) and crown-rump
length (CRL) during the first trimester
of pregnancy are already used effi-
ciently to detect chromosomal abnor-
malities.
1
This method can detect be-
tween 70% and 90% of Down’s
syndrome cases, with a false-positive
rate of 5%, depending on whether this
risk estimation is combined solely with
the risk because of maternal age or also
with first- or second-trimester mater-
nal serum markers.
2-5
However, the performance of this
technique depends largely on the preci-
sion of ultrasound measurement. Since
1997, the Royal College of Obstetricians
and Gynaecologists has recommended
the use of quality control when measur-
ing nuchal translucency.
6
Nicolaides and
the Fetal Medicine Foundation
7
(FMF)
defined methodological criteria for mea-
suring NT. They set up a clinical audit
based on analyzing 10 images selected by
the operator and also on all the NT and
CRL measurements to determine
whether the median score of each oper-
ator was close to the reference median.
Working on the criteria established by
the FMF, new image-scoring methods
(ISMs) were then defined. Herman et al
8
developed an ISM based on 6 criteria for
analyzing ultrasound images for NT
measurement. A total score for each im-
age could be obtained by adding together
the total scores of 3 major criteria
(scored 0 or 2) and 3 minor criteria
(scored 0 or 1). This ISM scores the NT
images but does not take into account
the CRL. A high level of interreviewer re-
liability between the scores obtained by 3
reviewers from the same hospital was ob-
served. However, the agreement of this
ISM was evaluated on only a 4-level cat-
egorization of the total score.
Wojdemann et al
9
developed an ISM
based on whether each of the 4 criteria
was acceptable. They found that the re-
producibility of the ISM was low and
concluded that such methods could not
be recommended for ongoing quality
control. Given that there has been no real
agreement on these ISMs, further studies
are necessary. Notably, the need to ana-
lyze CRL images and study these evalua-
tions criterion by criterion to improve
clinical practice and reviewer training
has become obvious.
The aims of this study were therefore:
(1) to describe a new ISM for analyzing
the NT and CRL images on ultrasound
scans; (2) to study interreviewer reliabil-
ity for this ISM on the total score and the
scores of each criterion before and after a
short training session; and (3) to esti-
Collège Français d’Echographie Foetale (Drs
Fries, Althuser, Fontanges, and Talmant)
Paris, France; Centre Pluridisciplinaire de
Diagnostic Prénatal (Dr Jouk), Centre
Hospitalier Universitaire de Grenoble,
Grenoble, France; Centre National de la
Recherche Scientifique (Ms Tindel and Dr
Duyme), Laboratoire de Biostatistique
Université Montpellier Nimes, Montpellier,
France.
Received May 13, 2006; revised July 23,
2006; accepted Oct. 11, 2006.
Reprints: Dr N. Fries, Les Tonnelles, 131
Avenue de Lodève, Montpellier, 34080,
France; fries@cfef.org.
0002-9378/$32.00
© 2007 Mosby, Inc. All rights reserved.
doi: 10.1016/j.ajog.2006.10.866
Research www. AJOG.org
272.e1 American Journal of Obstetrics & Gynecology MARCH 2007