+ Models

HANSUR-68; No. of Pages 8

Available online at

ScienceDirect
www.sciencedirect.com

Hand Surgery and Rehabilitation xxx (2016) xxx–xxx

Original article

Reproducibility of X-rays and CT arthrography in SLAC, SNAC,

SCAC wrists
Reproductibilité des radiographies et de l’arthroscanner de poignet dans les lésions SLAC,
SNAC et SCAC
R. Belhaouane a,*, N. Lebeau a, C. Maes-Clavier a, C. Hustin a, E. Krief a,b, B. Bonnaire c,
M. Warin c, V. Rotari a, E. David a
a
CHU d’Amiens, Orthopedic and Traumatologic Department, site sud, 80000 Amiens, France
b
CHU Ambroise-Paré, Orthopedic and Traumatologic Department, 75010 Paris, France
c
CHU d’Amiens, Radiologic department, site sud, 80000 Amiens, France
Received 27 January 2016; received in revised form 10 June 2016; accepted 6 August 2016

Abstract
The purpose of this study was to assess the inter-observer and intra-observer reproducibility of the interpretation of CT arthrography and plain
X-rays for scapholunate advanced collapse (SLAC), scaphoid non-union advanced collapse (SNAC) and scaphoid chondrocalcinosis advanced
collapse (SCAC) wrist conditions, as well as the clinical relevance of these imaging modalities. The CT and X-rays images were reviewed twice in a
blinded and randomized manner by two experienced orthopedic surgeons specialized in hand surgery, two orthopedic surgery residents and two
experienced radiologists specialized in bone and joint imaging. Cohen’s kappa and Fleiss’ kappa coefficients were used to analyze the
reproducibility of interpretation of the radiological examinations. With CT arthrography, the overall diagnosis was often a problem, in terms
of both inter- or intra-observer reproducibility. The assessment of the joint line appeared to be fairly reproducible for each observer but was poorly
reproducible between different observers. Plain X-rays are not sufficient to assess cartilage quality in degenerative wrist disease. CT arthrography is
a reliable examination, but its interpretation is not always standardized. Diagnostic arthroscopy may be justified in doubtful cases.
# 2016 SFCM. Published by Elsevier Masson SAS. All rights reserved.

Keywords: SLAC; SNAC; SCAC; Osteoarthritis; Wrist; Reproducibility; X-ray; CT arthrography

Résumé
Le but de cette étude était d’analyser la reproductibilité intra- et inter-observateur des interprétations de radiographies standards et
d’arthroscanner de poignet dans les SLAC wrist, SNAC wrist et SCAC wrist. Nous avons soumis en double lecture anonyme et randomisée
des radiographies standards et des arthroscanners de poignet à deux chirurgiens orthopédistes spécialisés de la chirurgie de la main, deux internes
de chirurgie orthopédique et deux radiologues confirmés spécialisés en imagerie ostéoarticulaire. Nous avons utilisé les coefficients Kappa de
Cohen et Kappa de Fleiss pour l’interprétation de la reproductibilité intra- et inter-observateur de chaque examen. Concernant l’interprétation des
arthroscanners, le diagnostic global a souvent posé problème, que ce soit en inter- ou en intra-observateur. Pour les deux examens, l’interprétation
des interlignes articulaires paraissait assez reproductible pour chaque observateur mais était peu reproductible entre les différents observateurs. Les
radiographies standard ne suffisent pas à évaluer la qualité du cartilage dans le cadre des pathologies dégénératives de poignet. L’arthro-TDM est
un bon examen, dont l’interprétation n’est pas toujours standardisée. En cas de doute, le recours à l’arthroscopie diagnostique paraît justifié.
# 2016 SFCM. Publié par Elsevier Masson SAS. Tous droits réservés.

Mots clés : SLAC ; SNAC ; SCAC ; Arthrose ; Poignet ; Reproductibilité ; Radiographie ; Arthroscanner

* Corresponding author. CHU d’Amiens Sud, Service de Chirurgie Orthopédique, 80000 Amiens, France.
E-mail address: ramy.b@hotmail.fr (R. Belhaouane).

http://dx.doi.org/10.1016/j.hansur.2016.08.006
2468-1229/# 2016 SFCM. Published by Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006
+ Models
HANSUR-68; No. of Pages 8

2 R. Belhaouane et al. / Hand Surgery and Rehabilitation xxx (2016) xxx–xxx

1. Introduction 2.4. Methods

Considerable progress has been made in our knowledge Imaging examinations were reviewed by two experienced
of degenerative arthritis of the wrist over the last decade. orthopedic surgeons specialized in hand surgery, two ortho-
Many studies have been conducted on this subject, especially pedic surgery residents and two experienced radiologists
concerning the therapeutic management and surgical specialized in bone and joint imaging.
options that can be proposed to patients depending on the Each observer did two blinded readings of each examination
degree of joint damage [1]. The preoperative diagnostic in the following order: examination of the series of plain
assessment must include imaging to define the lesion stage anteroposterior and lateral X-rays followed by examination of
[2–4]. the CT arthrography series. Each series was presented in
This study focused on degenerative arthritis of the wrist in random order. A second randomized review was performed
the context of scapholunate advanced collapse (SLAC) [5], 2 weeks later. Examinations were presented anonymously to
scaphoid non-union advanced collapse (SNAC) [6] and the observers: no information about the patient’s identity,
scaphoid chondrocalcinosis advanced collapse (SCAC) condi- demographics or clinical context was provided.
tions [7,8]. The various observers were asked to note the presence or
In our routine clinical practice, the imaging assessment absence of arthritic damage in the styloscaphoid (SS),
comprises plain X-rays combined with CT arthrography of the radioscaphoid (RS), radiolunate (RL), capitolunate (CL),
wrist. Inter-observer differences in the interpretation of these scaphotrapeziotrapezoid (STT), lunohamate (LH) (in case of
images have been observed. Viegas type 2 lunate), trapeziometacarpal (TM) and distal
The purpose of this study was to assess the inter-observer radioulnar (DRU) joints, in order to classify the lunate
and intra-observer reproducibility of the interpretation of CT according to the Viegas classification based on the presence
arthrography and plain X-rays for SLAC wrist, SNAC wrist and or absence of the lunohamate facet, the presence of ulnocarpal
SCAC wrist, as well as the clinical relevance of these imaging conflict and finally the stage of degenerative lesions of the wrist.
modalities. In 1984, Watson and Ballet [2] described a three-stage
classification for SLAC lesions that did not include radiolunate
2. Material and methods joint lesions. The SNAC classification described by Vender
et al. [4] also consisted of three stages. Several authors have
2.1. Data source subsequently described a fourth stage for SLAC wrist [1,9] and
SNAC wrist [10]. We therefore adopted a four-stage
We collected plain X-rays and CT arthrographies of the wrist classification of SLAC and SNAC lesions [1,11] (Fig. 1).
performed in our hospital’s radiology department between The following classification was used for SCAC wrists [8]:
January 2010 and January 2014. All patients with radiocarpal
arthritis and/or midcarpal arthritis secondary to SLAC, SNAC  SCAC 1: isolated involvement of radioscaphoid joint with
or SCAC diagnosed radiographically and/or by CT were verticalization of scaphoid;
included. Cases with incomplete imaging records or an  SCAC 2: early signs of scaphoid impaction into radius;
uncertain diagnosis were excluded. lunocapitate involvement;
 SCAC 3: scaphoid is completely impacted into radius;
2.2. Study population midcarpal dislocation with dorsal intercalated segmental
instability (DISI);
Fifty wrists (31 right and 19 left) from 48 patients were  SCAC 4: destruction of all carpal joints with involvement of
included, corresponding to 37 male and 11 female patients with radiolunate joint and signs of intercarpal fusion.
a mean age of 50 years (range 19–79 years). The number of
wrists per diagnosis and lesion stage was as follows: Each study observer was given a table to be completed for
each review.
 SNAC wrist: 15 wrists (5 stages 1, 2 stages 2, 7 stages 3, To analyze the reproducibility of the radiological inter-
1 stage 4); pretation, Cohen’s kappa coefficient was used to compare two
 SLAC wrist: 31 wrists (9 stages 1, 4 stages 2, 14 stages 3, observations by the same observer (intra-observer reproduci-
4 stage 4); bility) and to compare results for the same examination
 SCAC wrist: 4 wrists (2 stages 2, 1 stage 3, 1 stage 4). between two observers (inter-observer reproducibility). Fleiss’
kappa coefficient was used to assess inter-observer reprodu-
2.3. Ethical considerations cibility by multivariate analysis. The 95% confidence interval
(CI) for the overall kappa value was estimated [12–14]. Kappa
This study was part of a non-interventional research coefficients were interpreted according to the modalities
protocol. We obtained permission from the French data proposed by Landis and Koch as follows:
protection authority (CNIL authorization number:
mpQ1024233t) and ethical approval from our hospital’s  < 0: poor agreement;
institutional review board (CPP number: 00009118).  0.01–0.20: slight agreement;

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006
+ Models
HANSUR-68; No. of Pages 8

R. Belhaouane et al. / Hand Surgery and Rehabilitation xxx (2016) xxx–xxx 3

Fig. 1. Classification SLAC and SNAC wrist lesion stages (reproduced with the kind permission of Dr. Imbriglia).

 0.21–0.40: fair agreement; Intra-observer reproducibility with plain X-rays for the
 0.41–0.60: moderate agreement; overall diagnosis was substantial and perfect for residents and
 0.61–0.80: substantial agreement; radiologists and fair for the surgeons.
 0.81–1.00: perfect agreement.
3.2. Intra-observer reproducibility between 1st and 2nd
The P-value was deemed significant when less than 0.05. readings of CT arthrographies
The statistical analysis was performed by our Clinical Research
and Innovation Department. Intra-observer reproducibility for CT arthrographies
varied considerably by observer, with substantial and perfect
3. Results reproducibility for the residents for all joints except for
ulnocarpal conflict, which had moderate reproducibility
3.1. Intra-observer reproducibility between 1st and 2nd (Table 2).
readings of plain X-rays Intra-observer reproducibility was substantial and perfect
for the majority of joints for Surgeon 2 except for the DRU joint
Intra-observer reproducibility of plain X-rays varied by and ulnocarpal conflict (moderate), while the results were much
observer, with substantial and perfect reproducibility for the more widely scattered for Surgeon 1.
residents for all joints except ulnocarpal conflict, which had Substantial and perfect reproducibility were observed for
moderate reproducibility (Table 1). Radiologist 2, while reproducibility for Radiologist 1 was
Intra-observer reproducibility for the surgeons was sub- moderate for the STT and DRU joints and ulnocarpal conflict,
stantial and perfect for the majority of joints. It was moderate and substantial or perfect for all other joints.
for the radiolunate joint for Surgeon 1, and for the Intra-observer reproducibility for the CT arthrography
styloscaphoid, radiolunate, STT joints and Viegas classification diagnosis was substantial and perfect for Resident 1, Surgeon
for Surgeon 2. Intra-observer reproducibility for the DRU joint 2, and Radiologist 2, and moderate for the other observers.
was fair for Surgeon 2.
Intra-observer reproducibility was also substantial 3.3. Intra-observer reproducibility between 2nd reading of
and perfect for the radiologists for the majority of joints plain X-rays and 2nd reading of CT arthrographies
except for the radiolunate and DRU joints for Radiologist
1 and the STT joint and ulnocarpal conflict for Radiologist Intra-observer reproducibility between plain X-rays and CT
2 (moderate). arthrographies was moderate and below the mean for the

Table 1
Kappa values for intra-observer reproducibility between 1st and 2nd readings of plain X-rays with 95% CI (* when P  0.05).
Resident 1 Resident 2 Surgeon 1 Surgeon 2 Radiologist 1 Radiologist 2
SS 1 0.93 [0.79–1] 0.65 [0.36–0.93] 0.45 [0.11–0.79] 0.83 [0.67–0.99] 0.83 [0.61–1]
RS 0.96 [0.88–1] 0.96 [0.88–1] 0.80 [0.63–0.96] 0.69 [0.50–0.88] 0.87 [0.73–1] 0.84 [0.69–0.99]
RL 1 0.93 [0.79–1] 0.42 [0.11–0.74] 0.44 [0.08–0.80] 0.49 [0.12–0.87] 0.70 [0.38–1]
CL 1 0.96 [0.88–1] 0.75 [0.57–0.93] 0.71 [0.52–0.91] 0.87 [0.73–1] 1
STT 0.90 [0.70–1] 0.72 [0.53–0.91] 0.69 [0.45–0.92] 0.57 [0.36–0.78] 0.68 [0.48–0.88] 0.56 [0.33–0.79]
LH 0.92 [0.76–1] 0.87 [0.70–1] 1 0.71 [0.41–1] 1 0.83 [0.62–1]
TM 0.74 [0.46–1] 0.84 [0.67–1] 0.78 [0.48–1] 0.72 [0.49–0.95] 0.81 [0.59–1] 0.75 [0.54–0.95]
DRU 0.91 [0.78–1] 0.60 [0.31–0.88] 0.73 [0.48–0.98] 0.27 [0.01–0.53] 0.49 [0.26–0.72] 0.65 [0.37–0.93]
Conflict 0.90 [0.70–1] 0.56 [0.21–0.90] 0.78 [0.49–1] 0.67 [0.37–0.97] 0.78 [0.49–1] 0.56 [0.25–0.87]
Viegas 0.96 [0.88–1] 0.88 [0.73–1] 0.85 [0.65–1] 0.52 [0.28–0.75] 0.62 [0.39–0.86] 0.80 [0.63–0.96]
Diagnosis 0.85 0.89 0.33 0.28 0.75 0.68
SS: styloscaphoid; RS: radioscaphoid; RL: radiolunate; CL: capitolunate; STT: scaphotrapeziotrapezoid; LH: lunohamate (Viegas type 2); TM: trapeziometacarpal;
DRU: distal radioulnar.

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006
+ Models
HANSUR-68; No. of Pages 8

4 R. Belhaouane et al. / Hand Surgery and Rehabilitation xxx (2016) xxx–xxx

Table 2
Kappa values for intra-observer reproducibility between 1st and 2nd readings of CT arthrographies with 95% CI (* when P  0.05).
Resident 1 Resident 2 Surgeon 1 Surgeon 2 Radiologist 1 Radiologist 2
SS 1 0.90 [0.77–1] 0.48 [–0.14–1] 0.79 [0.56–1] 0.90 [0.75–1] 1
RS 1 0.75 [0.57–0.94] 0.74 [0.55–0.93] 0.92 [0.81–1] 0.64 [0.43–0.84] 0.95 [0.85–1]
RL 1 0.65 [0.41–0.88] 0.52 [0.25–0.79] 0.88 [0.64–1] 0.71 [0.51–0.90] 0.68 [0.47–0.88]
CL 0.96 [0.88–1] 0.68 [0.49–0.88] 0.63 [0.41–0.85] 0.83 [0.66–0.99] 0.67 [0.47–0.87] 0.82 [0.66–0.99]
STT 0.91 [0.74–1] 0.70 [0.49–0.90] 0.48 [0.23–0.73] 0.69 [0.49–0.90] 0.50 [0.27–0.72] 0.83 [0.64–1]
LH 0.87 [0.70–1] 0.86 [0.70–1] 0.68 [0.26–1] 0.84 [0.68–1] 0.73 [0.54–0.93] 0.85 [0.64–1]
TM 0.92 [0.77–1] 0.73 [0.45–1] 0.50 [0.15–0.84] 0.67 [0.42–0.91] 0.66 [0.46–0.87] 0.84 [0.69–0.99]
DRU 0.95 [0.85–1] 0.39 [0.05–0.74] 0.29 [0.02–0.56] 0.52 [0.29–0.75] 0.48 [0.24–0.71] 0.72 [0.53–0.91]
Conflict 0.92 [0.77–1] 0.62 [0.35–0.90] 0.85 [0.55–1] 0.49 [0.21–0.78] 0.50 [0.22–0.77] 0.72 [0.52–0.93]
Viegas 1 0.88 [0.64–1] 0.61 [0.38–0.84] 0.81 [0.60–1] 0.79 [0.60–0.99] 1
Diagnosis 0.98 0.43 0.47 0.67 0.49 0.75
SS: styloscaphoid; RS: radioscaphoid; RL: radiolunate; CL: capitolunate; STT: scaphotrapeziotrapezoid; LH: lunohamate (Viegas type 2); TM: trapeziometacarpal;
DRU: distal radioulnar.

Table 3
Kappa values for intra-observer reproducibility between 2nd reading of plain X-rays and 2nd reading of CT arthrographies with 95% CI.
Resident 1 Resident 2 Surgeon 1 Surgeon 2 Radiologist 1 Radiologist 2
SS 0.83 [0.65–1] 0.54 [0.28–0.81] 0.07 [ 0.15–0.01]a 0.65 [0.37–0.93] 0.68 [0.47–0.89] 0.41 [0.01–0.80]
RS 0.80 [0.63–0.96] 0.75 [0.56–0.94] 0.47 [0.24–0.71] 0.76 [0.58–0.94] 0.50 [0.29–0.72] 0.56 [0.35–0.77]
RL 0.77 [0.52–1] 0.51 [0.24–0.77] 0.12 [ 0.17–0.42]a 0.39 [ 0.04–0.82] 0.17 [0.01–0.33] 0.21 [0.06–0.36]
CL 0.92 [0.80–1] 0.52 [0.29–0.76] 0.76 [0.57–0.94] 0.71 [0.51–0.90] 0.68 [0.48–0.88] 0.54 [0.34–0.74]
STT 0.62 [0.28–0.96] 0.40 [0.14–0.65] 0.18 [ 0.11–0.46]a 0.25 [ 0.02–0.51]a 0.36 [0.10–0.62] 0.38 [0.16–0.60]
LH 0.70 [0.44–0.96] 0.56 [0.30–0.81] 1 0.41 [0.11–0.72] 0.38 [0.08–0.67] 0.13 [ 0.02–0.27]a
TM 0.70 [0.43–0.98] 0.60 [0.31–0.88] 0.25 [ 0.11–0.61]a 0.73 [0.51–0.95] 0.37 [0.15–0.59] 0.41 [0.18–0.64]
DRU 0.60 [0.36–0.85] 0.12 [ 0.34–0.10]a 0.41 [0.16–0.66] 0.42 [0.12–0.73] 0.36 [0.11–0.62] 0.35 [0.16–0.54]
Conflict 0.62 [0.29–0.96] 0.22 [ 0.11–0.55]a 0.88 [0.64–1] 0.51 [0.21–0.82] 0.40 [0.12–0.68] 0.63 [0.39–0.87]
Viegas 0.84 [0.69–0.99] 0.42 [0.11–0.74] 0.38 [0.14–0.62] 0.34 [0.14–0.54] 0.22 [0.08–0.36] 0.28 [0.05–0.50]
Diagnosis 0.73 0.29 0.22 0.36 0.30 0.43
SS: styloscaphoid; RS: radioscaphoid; RL: radiolunate; CL: capitolunate; STT: scaphotrapeziotrapezoid; LH: lunohamate (Viegas type 2); TM: trapeziometacarpal;
DRU: distal radioulnar.
a
P  0.05

majority of joints except for Resident 1 (substantial and perfect) series of examinations in random order with a minimum
(Table 3). interval of 2 weeks between each reading to avoid a possible
Intra-observer reproducibility between plain X-rays and CT memory effect between the 1st and 2nd readings and between
arthrographies was below the mean for the Viegas classification plain X-rays versus CT arthrography. A classification grid was
except for Resident 1 (substantial). provided to standardize interpretation, which allowed us to
compare the results.
3.4. Inter-observer reproducibility of 2nd reading of plain Several joints were studied in order to provide an overall
X-rays and CT arthrographies between observer pairs assessment of the wrist. We studied not only the joints involved
in SNAC wrist, SLAC wrist and SCAC wrist, but also other
Inter-observer reproducibility of plain X-rays and CT joints that can be involved concurrently. This concept must be
arthrographies between observer pairs was moderate and integrated into the management of degenerative wrist disease.
below the mean for interpretation of joints and for the final For example, in the case of type 2 and 3 SNAC wrist,
diagnosis, but with slightly better results for the pair of trapeziometacarpal joint involvement must be taken into
radiologists (Table 4). account when deciding on surgery.
As previously proposed by Chammas’ team [15], we suggest
3.5. Inter-observer reproducibility of 1st and 2nd readings that simultaneous involvement of other joints should be added
of plain X-rays and CT arthrographies to the classification of SLAC, SNAC and SCAC wrist by
indicating ‘‘+’’ or ‘‘ ’’ in front of each joint studied (e.g.:
Inter-observer reproducibility of plain X-rays and CT SLAC wrist stage 2 without STT osteoarthritis, with
arthrographies was moderate and fair (Table 5). trapeziometacarpal osteoarthritis, lunate with uninjured luno-
hamate facet, without DRU osteoarthritis: SLAC 2 STT TM+
4. Discussion Viegas 2 LH DRU ).
Other studies have assessed the reproducibility of pre-
In this study, all imaging examinations were made operative radiological examinations in routine clinical practice
anonymous prior to review. Each observer reviewed each [16–18], not only in the wrist but in other joints. The majority of

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006
+ Models
HANSUR-68; No. of Pages 8

R. Belhaouane et al. / Hand Surgery and Rehabilitation xxx (2016) xxx–xxx 5

Table 4
Kappa values for inter-observer reproducibility of 2nd reading of plain X-rays and CT arthrographies between observer pairs with 95% CI.
X-rays CT arthrographies
Residents Surgeons Radiologists Residents Surgeons Radiologists
a
SS 0.29 [ 0.03–0.61] 0.45 [0.11–0.79] 0.42 [0.19–0.65] 0.07 [ 0.22–0.35] 0.29 [ 0.03–0.60] 0.23 [ 0.04–0.51]
RS 0.42 [0.19–0.65] 0.56 [0.34–0.79] 0.72 [0.53–0.91] 0.54 [0.31–0.78] 0.61 [0.41–0.81] 0.69 [0.48–0.90]
RL 0.34 [ 0.02–0.69] 0.17 [ 0.13–0.47]a 0.49 [0.12–0.87] 0.14 [ 0.14–0.42]a 0.35 [0.06–0.64] 0.48 [0.25–0.72]
CL 0.52 [0.28–0.75] 0.72 [0.54–0.91] 0.79 [0.62–0.96] 0.52 [0.30–0.75] 0.48 [0.28–0.68] 0.56 [0.35–0.77]
STT 0.21 [0.04–0.38] 0.19 [ 0.06–0.45]a 0.56 [0.33–0.79] 0.19 [ 0.04–0.41]a 0.54 [0.29–0.79] 0.43 [0.23–0.64]
LH 0.64 [0.32–0.96] 1 0.76 [0.44–1] 0.40 [0.07–0.73] 0.79 [0.52–1] 0.57 [0.31–0.83]
TM 0.26 [ 0.05–0.57] 0.43 [0.13–0.73] 0.63 [0.39–0.87] 0.29 [ 0.06–0.64] 0.38 [0.07–0.68] 0.72 [0.53–0.91]
DRU 0.28 [ 0.01–0.57] 0.45 [0.11–0.79] 0.31 [0.08–0.55] 0.04 [ 0.30–0.22]a 0.57 [0.32–0.81] 0.72 [0.53–0.91]
Conflict 0.25 [ 0.13–0.62]a 0.62 [0.28–0.96] 0.41 [0.08–0.74] 0.41 [0.08–0.74] 0.35 [0.02–0.69] 0.56 [0.31–0.82]
Viegas 0.41 [0.19–0.64] 0.30 [0.11–0.48] 0.33 [0.13–0.53] 0.21 [ 0.01–0.43] 0.21 [0.06–0.35] 0.63 [0.38–0.89]
Diagnosis 0.25 0.24 0.37 0.18 0.30 0.48
SS: styloscaphoid; RS: radioscaphoid; RL: radiolunate; CL: capitolunate; STT: scaphotrapeziotrapezoid; LH: lunohamate (Viegas type 2); TM: trapeziometacarpal;
DRU: distal radioulnar.
a
P  0.05

Table 5
Kappa values for inter-observer reproducibility of 1st and 2nd readings of plain X-rays and CT arthrographies.
X-rays CT arthrographies
1st reading 2nd reading 1st reading 2nd reading
SS 0.36 0.46 0.30 0.27
RS 0.61 0.64 0.62 0.62
RL 0.37 0.26 0.23 0.28
CL 0.57 0.66 0.61 0.57
STT 0.34 0.28 0.30 0.28
LH 0.81 0.88 0.37 0.27
TM 0.51 0.45 0.42 0.39
DRU 0.36 0.29 0.20 0.31
Conflict 0.42 0.42 0.32 0.42
Viegas 0.30 0.32 0.31 0.35
Diagnosis 0.31 0.34 0.31 0.33
SS: styloscaphoid; RS: radioscaphoid; RL: radiolunate; CL: capitolunate; STT: scaphotrapeziotrapezoid; LH: lunohamate (Viegas type 2); TM: trapeziometacarpal;
DRU: distal radioulnar.

these studies used the Kappa coefficient, which confirms the et al. [29] studied CT arthrography interpretation by using MR
validity of our approach. This statistical method has the arthrography as the gold standard and reported sensitivity and
advantage of being relevant, simple and easy to interpret. specificity of 70% and 93%, respectively, for the detection of
Statistically significant results were obtained with this method radioscaphoid chondral lesions and 66% and 92%, respectively,
despite the relatively small number of cases studied. for the radiolunate joint. Inter-observer reproducibility of
Several studies have demonstrated that CT arthrography is chondral and ligament injuries was very poor.
superior to plain X-rays in terms of sensitivity and specificity Another study comparing MR arthrography of the wrist
[18,19]. CT arthrography provides more precise and more versus arthroscopy as the gold standard showed that MR
complete analysis of wrist lesions and can show bone, cartilage arthrography was an excellent examination for the diagnosis of
and ligament lesions. It allows assessment of bone stock and ligament injuries, but reported more cautious conclusions for
other wrist diseases that could subsequently become decom- the diagnosis of cartilage lesions [30].
pensated. Plain X-rays and MRI were compared to unenhanced CT for
Various studies have proposed treatment strategies or the assessment of bone erosion in rheumatoid arthritis wrists
surgical procedures according to the lesion stage [20–25]. and found moderate sensitivity, high specificity and almost
Radial styloidectomy [26] is usually proposed in stage 1 perfect reproducibility of MRI [31].
disease, proximal row carpectomy in stage 2, midcarpal Vishwanathan et al. [32] reported poor reproducibility of the
arthrodesis in stage 3 [2], and total arthrodesis or total wrist SNAC and SLAC wrist classification based on interpretation of
arthroplasty [27] in stage 4. Total denervation [28] is another plain X-rays, with modest results for both intra-observer and
surgical treatment option that can be used regardless of disease inter-observer reproducibility.
stage. Compared with radiological examinations, it is certain that
Other authors have compared the performance of CT arthroscopy is more appropriate as it can provide a clear and
arthrography with that of MRI or MR arthrography. Koskinen precise diagnosis. Wrist arthroscopy has taken an important

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006
+ Models
HANSUR-68; No. of Pages 8

6 R. Belhaouane et al. / Hand Surgery and Rehabilitation xxx (2016) xxx–xxx

Fig. 2. Plain X-ray (A) and CT arthrography (B) of a wrist with SNAC 1 SC+ lesion.

place in the therapeutic strategy and the diagnostic approach in understand these differences in interpretation. One of
traumatic pathologies of the wrist. Arthroscopic criteria for the causes of conflicting results could be the lack of
dating wrist sprains have been proposed including a distinction in the type of cartilage lesions (height, depth)
classification of cartilage lesions [33]. during the joint analysis. For example, a partial lesion could
In the present study, the overall diagnosis based on be considered pathological by some observers but not by
interpretation of CT arthrographies was often difficult, as others.
evidenced by both the intra-observer and inter-observer We also encountered several cases of isolated scaphocapitate
reproducibility values. Joint interpretation on the two lesion with no lunocapitate involvement (Fig. 2). This type of
examinations was fairly reproducible for each observer, but lesion was difficult to classify because some participants
was poorly reproducible between the various observers. A distinguished it from the midcarpal lesion, while other
slightly better reproducibility was observed for observer pairs. participants considered it to be midcarpal osteoarthritis. This
The comparison between plain X-rays and CT arthrography situation was essentially encountered in stage 1 and 2 SNAC
revealed poor reproducibility of the joint interpretation and the lesions associated with scaphocapitate involvement. We
overall diagnosis. No particular disease or stage of disease was propose that this lesion should be distinguished from midcarpal
associated with poorer or better reproducibility. arthritis and should be scored separately; i.e. in stage 1 and 2,
Observers were asked about any difficulties encountered indicate the presence or absence of scaphocapitate involvement
during reading of these examinations in order to more clearly by SC+ or SC .

Fig. 3. Plain X-ray (A) and CT arthrography (B) of a SLAC wrist with styloscaphoid and lunocapitate osteoarthritis and no involvement of the radioscaphoid joint.

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006
+ Models
HANSUR-68; No. of Pages 8

R. Belhaouane et al. / Hand Surgery and Rehabilitation xxx (2016) xxx–xxx 7

Other cases of SNAC and SLAC may also be difficult to Disclosure of interest
classify, such as concomitant styloscaphoid and lunocapitate
lesions with no radioscaphoid involvement (Fig. 3), which The authors declare that they have no competing interest.
raises a diagnostic problem between stage 1 with lunocapitate
osteoarthritis and stage 3. This configuration was not reported Acknowledgements
by Watson in his description of the course of degenerative Our hospital’s Clinical Research and Innovation Department
lesions in SLAC and SNAC wrists. Prospective long-term for statistical analysis.
follow-up of degenerative wrists could determine the clinical
relevance of these unusual cases. References
The radiological images were presented to the observers
anonymously, with no information on the history of the disease, [1] Weiss KE, Rodner CM. Osteoarthritis of the wrist. J Hand Surg Am
the patient’s surgical or traumatic history, the presence of 2007;32:725–46.
inflammatory joint disease, or clinical findings. In routine [2] Watson HK, Ballet FL. The SLAC wrist: scapholunate advanced collapse
clinical practice, these data are available at the time of pattern of degenerative arthritis. J Hand Surg Am 1984;9:358–65.
[3] Watson HK, Brenner LH. Degenerative disorders of the wrist. J Hand Surg
interpretation and can help to guide the diagnosis. The Am 1985;10:1002–6.
examiner can also be assisted by comparing the various [4] Vender MI, Watson HK, Wiener BD, Black DM. Degenerative change in
examinations performed and reviewing any other older symptomatic scaphoid nonunion. J Hand Surg Am 1987;12:514–9.
examinations in order to assess possible deterioration. [5] Krakauer JD, Bishop AT, Cooney WP. Surgical treatment of scapholunate
advanced collapse. J Hand Surg Am 1994;19:751–9.
In some cases, degenerative lesions can be secondary to
[6] Inoue G, Sakuma M. The natural history of scaphoid non-union. Radio-
several different diseases, for example a combination of partial graphical and clinical analysis in 102 cases. Arch Orthop Trauma Surg
scapholunate ligament tear with collapse secondary to malunion 1996;115:1–4.
of the distal radius. These more complex diagnoses may have [7] Bardin T, Fritz P. [Microcrystal deposit arthropathies of the wrist]. Ann
been a problem for the observers of this study, but also occur in Radiol (Paris) 1992;35:402–6.
clinical practice, where they raise the same diagnostic problems. [8] Romano S. Arthrose non traumatique du poignet: la chondrocalcinose.
Chir Main 2003;22:285–92.
Another possible source of error could be technical, related [9] Peterson B, Szabo RM. Carpal osteoarthrosis. Hand Clin 2006;22:517–28
to difficulties during contrast agent injection, particularly in the [abstract vii].
presence of marked wrist stiffness. Absence of sufficient [10] Trumble TE, Salas P, Barthel T, Robert 3rd KQ. Management of scaphoid
diffusion of contrast agent can be a source of error. nonunions. J Am Acad Orthop Surg 2003;11:380–91.
Finally, the results of this study have direct implications in [11] Imbriglia JE. Proximal row carpectomy. In: Berger RA, Weiss APC,
editors. Hand surgery. Philadelphia: Lippincott, Williams & Wilkins;
the field of scientific research and evaluation of surgical 2004. p. 1332–4.
practice. Many scientific publications comparing the efficacy of [12] Landis JR, Koch GG. The measurement of observer agreement for
two procedures have been based on homogeneous patient categorical data. Biometrics 1977;33:159–74.
groups with a well-defined diagnosis and classification. [13] Kundel HL, Polansky M. Measurement of observer agreement. Radiology
2003;228:303–8.
Although the conclusions derived from these studies apply
[14] Petrie A. Statistics in orthopaedic papers. J Bone Joint Surg Br 2006;88:
to a particular stage of disease, they may no longer be valid if 1121–36.
the initial diagnosis is not reliable. Wrist arthroscopy may [15] Lenoir H, Toffoli A, Coulet B, Lazerges C, Waitzenegger T, Chammas M.
provide a more reliable diagnosis for difficult cases. Radiocapitate congruency as a predictive factor for the results of proximal
row carpectomy. J Hand Surg Am 2015;40:1088–94.
5. Conclusion [16] Lecouvet FE, Dorzée B, Dubuc JE, Berg BCV, Jamart J, Malghem J.
Cartilage lesions of the glenohumeral joint: diagnostic effectiveness of
multidetector spiral CT arthrography and comparison with arthroscopy.
Plain X-rays and CT arthrography are essential examinations Eur Radiol 2007;17:1763–71.
in our diagnostic approach. Plain X-rays are not sufficient to [17] De Filippo M, Bertellini A, Pogliacomi F, Sverzellati N, Corradi D,
assess cartilage quality in degenerative wrist disease. CT Garlaschi G, et al. Multidetector computed tomography arthrography
of the knee: diagnostic accuracy and indications. Eur J Radiol 2009;
arthrography is a reliable examination, but its interpretation is
70:342–51.
not always standardized. Good quality contrast agent injection [18] De Filippo M, Pogliacomi F, Bertellini A, Araoz PA, Averna R, Sverzellati
is required for detailed analysis of cartilage. The radiologist N, et al. MDCT arthrography of the wrist: diagnostic accuracy and
must carefully examine the joint in all planes and must specify indications. Eur J Radiol 2010;74:221–5.
the type of lesion (thinning of the cartilage, ulceration and [19] Crema MD, Zentner J, Guermazi A, Jomaah N, Marra MD, Roemer FW.
dimensions). Diagnostic arthroscopy may be justified in Scapholunate advanced collapse and scaphoid nonunion advanced collapse:
MDCT arthrography features. AJR Am J Roentgenol 2012;199:W202–7.
doubtful cases. The reliability of this examination could be [20] Merrell GA, Weiss APC. What is the best treatment for early degenerative
clearly established through a study comparing CT arthrography osteoarthritis of the wrist? In: Wright JG, editor. Evidence-based ortho-
and surgical findings. paedics: the best answers to clinical questions. Philadelphia: Saunders
Elsevier; 2009. p. 84–6.
Funding statement [21] Kiefhaber TR. Management of scapholunate advanced collapse pattern of
degenerative arthritis of the wrist. J Hand Surg Am 2009;34:1527–30.
[22] Strauch RJ. Scapholunate advanced collapse and scaphoid nonunion
The authors received no financial support for the research, advanced collapse arthritis-update on evaluation and treatment. J Hand
authorship, and/or publication of this article. Surg 2011;36:729–35.

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006
+ Models
HANSUR-68; No. of Pages 8

8 R. Belhaouane et al. / Hand Surgery and Rehabilitation xxx (2016) xxx–xxx

[23] Le Nen D, Richou J, Simon E, Le Bourg M, Nabil N, de Bodman C, et al. [29] Koskinen SK, Haapamäki VV, Salo J, Lindfors NC, Kortesniemi M,
The arthritic wrist. I˘the degenerative wrist: surgical treatment approa- Seppälä L, et al. CT arthrography of the wrist using a novel, mobile,
ches. Orthop Traumatol Surg Res 2011;97:S31–6. dedicated extremity cone-beam CT (CBCT). Skeletal Radiol
[24] Laulan J, Bacle G, de Bodman C, Najihi N, Richou J, Simon E, et al. The 2013;42:649–57.
arthritic wrist. II-The degenerative wrist: indications for different surgical [30] Schmitt R, Christopoulos G, Meier R, Coblenz G, Fröhner S, Lanz U, et al.
treatments. Orthop Traumatol Surg Res 2011;97:S37–41. [Direct MR arthrography of the wrist in comparison with arthroscopy: a
[25] Shah CM, Stern PJ. Scapholunate advanced collapse (SLAC) and scaph- prospective study on 125 patients]. RöFo 2003;175:911–9.
oid nonunion advanced collapse (SNAC) wrist arthritis. Curr Rev Mus- [31] Døhn UM, Ejbjerg BJ, Hasselquist M, Narvestad E, Moller J, Thomsen
culoskelet Med 2013;6:9–17. HS, et al. Detection of bone erosions in rheumatoid arthritis wrist joints
[26] Nakamura T, Cooney 3rd WP, Lui WH, Haugstvedt JR, Zhao KD, with magnetic resonance imaging, computed tomography and radiogra-
Berglund L, et al. Radial styloidectomy: a biomechanical study on phy. Arthritis Res Ther 2008;10:R25.
stability of the wrist joint. J Hand Surg 2001;26:85–93. [32] Vishwanathan K, Hearnden A, Talwalkar S, Hayton M, Murali SR, Trail
[27] Yeoh D, Tourret L. Total wrist arthroplasty: a systematic review of the IA. Reproducibility of radiographic classification of scapholunate ad-
evidence from the last five years. J Hand Surg Eur Vol 2015;40:458–68. vanced collapse (SLAC) and scaphoid nonunion advanced collapse
[28] Foucher G, Da Silva JB, Ferreres A. [Total denervation of the wrist. (SNAC) wrist. J Hand Surg Eur Vol 2013;38:780–7.
Apropos of 50 cases]. Rev Chir Orthop Reparatrice Appar Mot [33] Van Overstraeten L, Camus E-J. Arthroscopic criteria for dating wrist
1992;78:186–90. sprains. Chir Main 2012;31:171–5.

Please cite this article in press as: Belhaouane R, et al. Reproducibility of X-rays and CT arthrography in SLAC, SNAC, SCAC wrists. Hand
Surg Rehab (2016), http://dx.doi.org/10.1016/j.hansur.2016.08.006