Radiographic scoring methods in hand osteoarthritis - a systematic literature search and descriptive review


Osteoarthritis and Cartilage 22 (2014) 1710e1723
Radiographic scoring methods in hand osteoarthritis e a systematic
literature search and descriptive review

A.W. Visser y *, P. Bøyesen z, I.K. Haugen z, J.W. Schoones x, D.M. van der Heijde y z,
F.R. Rosendaal k, M. Kloppenburg y k
y Department of Rheumatology, Leiden University Medical Center, Leiden, The Netherlands
z Department of Rheumatology, Diakonhjemmet Hospital, Oslo, Norway
x Walaeus Library, Leiden University Medical Center, Leiden, The Netherlands
k Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, The Netherlands
a r t i c l e i n f o

Article history:
Received 21 January 2014
Accepted 30 May 2014

Keywords:
Osteoarthritis
Hand
Radiography
Systematic review
* Address correspondence and reprint requests to: V
Medical Center, Department of Rheumatology, C1-R, P
The Netherlands. Tel: 31-71-5263265; Fax: 31-71-526

E-mail address: a.w.visser@lumc.nl (A.W. Visser).

http://dx.doi.org/10.1016/j.joca.2014.05.026
1063-4584/© 2014 Osteoarthritis Research Society In
s u m m a r y

Objective: This systematic literature review aimed to evaluate the use of conventional radiography (CR)
in hand osteoarthritis (OA) and to assess the metric properties of the different radiographic scoring
methods.
Design: Medical literature databases up to November 2013 were systematically reviewed for studies
reporting on radiographic scoring of structural damage in hand OA. The use and metric properties of the
scoring methods, including discrimination (reliability, sensitivity to change), feasibility and validity, were
evaluated.
Results: Of the 48 included studies, 10 provided data on reliability, 11 on sensitivity to change, four on
feasibility and 36 on validity of radiographic scoring methods. Thirteen different scoring methods have
been used in studies evaluating radiographic hand OA. The number of examined joints differed exten-
sively and the obtained scores were analyzed in various ways. The reliability of the assessed radiographic
scoring methods was good for all evaluated scoring methods, for both cross-sectional and longitudinal
radiographic scoring. The responsiveness to change was similar for all evaluated scoring methods. There
were no major differences in feasibility between the evaluated scoring methods, although the evidence
was limited. There was limited knowledge about the validity of radiographic OA findings compared with
clinical nodules and deformities, whereas there was better evidence for an association between radio-
graphic findings and symptoms and hand function.
Conclusions: Several radiographic scoring methods are used in hand OA literature. To enhance compa-
rability across studies in hand OA, consensus has to be reached on a preferred scoring method, the
examined joints and the used presentation of data.

© 2014 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.
Introduction metacarpaphalangeal (MCP) joints3. Currently, no structure modi-
Osteoarthritis (OA) is the most common musculoskeletal dis-
order, frequently affecting the hands1,2. Hand OA is characterized by
the formation of bony enlargements and deformities, most
frequently occurring in the distal interphalangeal (DIP) joints and
first carpometacarpal (CMC1) joints, less often in the proximal
interphalangeal (PIP) joints and least prevalent in
isser A.W., Leiden University
.O. Box 9600, 2300 RC Leiden,
6752.

ternational. Published by Elsevier L
fying treatments are available. To date, few high-quality clinical
trials have been performed in hand OA4,5. A key problem in the lack
of high-quality clinical trials in hand OA is the lack of standardi-
zation of outcome measures4,6. The Outcome Measures in Rheu-
matoid Clinical Trials (OMERACT) and Osteoarthritis Research
Society International (OARSI) Task Force on Clinical Trials Guide-
lines defined core domains to describe outcomes in clinical trials.
One of these domains for structure modifying trials was
imaging.7e9

Conventional radiography (CR) is commonly used to assess
structural damage in hand OA, as they are widely available and
relatively cheap. Radiography allows visualization of osteophytes,
joint space narrowing (JSN), subchondral cysts, sclerosis and cen-
tral erosions.
td. All rights reserved.

mailto:a.w.visser@lumc.nl
http://crossmark.crossref.org/dialog/?doi=10.1016/j.joca.2014.05.026&domain=pdf
http://dx.doi.org/10.1016/j.joca.2014.05.026
http://dx.doi.org/10.1016/j.joca.2014.05.026
http://dx.doi.org/10.1016/j.joca.2014.05.026


A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e1723 1711
Several standardized scoring methods are available such as the
KellgreneLawrence (KL)10, Kessler11 and Kallman grading scales12,
the OARSI scoring atlas13, the VerbruggeneVeys anatomical phase
score14, and the Gent University scoring system (GUSS)15. These
scores differ in the joints that are assessed, the type of scores
(composite score or individual feature scores), and the total score
ranges.

Most scoring methods have been shown to be reliable in-
struments for the assessment of structural damage in hand OA as
well as its change15e17. However, a systematic comparison of the
different scoring methods that will help to decide on a recom-
mended method has not been performed.

We performed a systematic review to evaluate the use of CR in
studies on hand OA and to assess the metric properties of the
different radiographic scoring methods18. To this end we made use
of the OMERACT filter19, focusing on aspects of discrimination
(reliability and sensitivity to change), feasibility and truth (validity)
of the radiographic scoring methods available in hand OA.
Methods

Identification of studies

Incooperationwitha medical librarian (JWS), a systemic literature
search was performed to obtain all manuscripts reporting on any
radiographic scoring methods assessing the nature, severity and
progression of structural damage in hand OA. Medical literature da-
tabases (PubMed, Embase, Web of Science, COCHRANE and CINAHL)
were searched up to November 2013, using all variations of the
following key words ‘hand’, ‘osteoarthritis’, ‘radiography’, ‘reli-
ability’, ‘validity’, ‘sensitive’ and ‘feasibility’ (see Supplementary File
For Exact Search Strings).
Inclusion and exclusion criteria

First all retrieved titles were screened, subsequently selected
abstracts were reviewed and finally full text articles of the
remaining references were read by one reviewer (AWV). A random
sample of 150 titles was also reviewed by a second reviewer (MK),
resulting in a similar selection of titles. In case of uncertainties in
the reviewing process by the single reviewer, these were discussed
and solved with MK. The metric properties of the studied radio-
graphic scoring methods were evaluated according to four items:
reliability, sensitivity to change, feasibility and validity. Inclusion
criteria required for studies to evaluate these items differed per
item:

- Reliability was evaluated in studies describing the reliability of
two or more scoring methods performed on the same radio-
graphs and by the same reader. Both cross-sectional and longi-
tudinal studies were included.

- Sensitivity to change was evaluated in longitudinal studies of at
least one year, in which hand OA was assessed by at least two
radiographic scoring methods. Studies with a follow-up dura-
tion between one and three years using only one radiographic
scoring method were also included.

- Feasibility was evaluated in studies describing the feasibility of
one or more scoring methods.

- Validity was evaluated in studies comparing a radiographic
scoring method with other measurements of structural damage
such as magnetic resonance imaging (MRI), computed tomog-
raphy (CT), ultrasound (US), digital photography, histology or
nodes at physical examination. In addition, validity was evalu-
ated in studies comparing radiographic findings to clinical signs
such as hand function or symptoms. Both cross-sectional and
longitudinal studies were included.

Studies that fulfilled the requirements for at least one of these
four items were included in this review.

Animal studies, reviews, abstracts, letters to the editor and
studies reporting on musculoskeletal diseases other than hand OA
or in languages other than English were excluded.

Data extraction

A standardized form was used to extract information about the
following data: (1) study population (population size, setting, age,
sex), (2) applied radiographic scoring methods, (3) performance of
the scoring (number of readers, consensus/independent reading,
(4) assessed joints, (5) level of analyses of obtained scores (joint,
joint group or patient level) and used definition of outcome (e.g.,
summed scores (total or per feature), counts of number of affected
joints, dichotomized outcome), (6) results concerning: reliability
(intraclass correlation coefficient (ICC), kappa-value, percentage of
agreement, smallest detectable change (SDC)), sensitivity to change
(percentage of change, amount of change, standardized response
mean (SRM)), feasibility (time needed to perform scoring), validity
(correlations, associations and measures of agreement between
radiographic scores and other measures). From a random number
of studies data were also extracted by MK and all extracted results
were discussed with MK.

Statistical analyses

Due to the heterogeneity of the studies and the difference in
outcome measures that were used it was not possible to perform a
meta-analysis. Therefore we chose to perform a descriptive review.

Results

Literature flow

After removing duplicate references, 1873 unique references
were identified [Fig. 1]. After reviewing 133 abstracts and 80 full-
text articles, 48 articles were included in this review. Of the
included studies, 10 fulfilled the inclusion criteria for evaluation of
reliability12,16,17,20e26, 11 for sensitivity to change14,16,17,24e31, four
for feasibility11,16,17,22, and 36 for validity of radiographic scoring
methods.20e24,32e62

Evaluation of radiographic scoring methods was the primary aim
in 10 of the included studies11,12,14,16,17,22,26,27,59,60. The other studies
used radiographic scoring to identify prevalence or progression of
radiographic OA features (n ¼ 7)20,25,28e30,33,34, or to compare ob-
tained scores with other outcome measures (other imaging methods,
clinical outcomes, histology) (n ¼ 31).21,23,24,31,32,35e38,40e58,61e63

The characteristics of the evaluated or applied radiographic
scoring methods (except for non-validated methods) are depicted
in Table I.

Study characteristics

The characteristics of the 48 included studies are depicted in
Table II. Most studies included more women than men and most of
the studied individuals were aged >50 years. As shown in Table II, a
wide variety of scoring methods (n ¼ 13) was used to assess
radiographic (signs of) hand OA. The KL scoring method was used
most frequently (n ¼ 24), followed by the OARSI scoring method
(n ¼ 18). Other scoring methods were the Kallman (n ¼ 9), indi-
vidual features following non-validated methods (n ¼ 7),



Fig. 1. Overview of literature research.

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e17231712
anatomical phases (n ¼ 6), anatomical lesions (n ¼ 2) and automatic
JSW measurement (n ¼ 3). The GUSS, Burnett, Kessler, Lane, Eaton
and a non-validated global score were all used in only one study.
Although the majority of studies used only one radiographic
scoring method, 15 studies used more than one method.

The examined joint groups differed between the studies: DIPs
and PIPs were assessed most frequently (in 48 and 46 studies,
respectively), followed by the CMC1s (n ¼ 34), MCPs (n ¼ 30), IP1s
(n ¼ 23) and the scaphotrapezotrapezoidal (STT) joints (n ¼ 8).

The way the analysis of the radiographic scores were executed
was quite different across the studies; (1) the score of one joint (the
most severely affected) from a joint group, hand or patient33,36,37,
43,46,50, (2) sum score for all joints and features14,16,17,20e22,24e26,31,
34,38,44,45, (3) sum scores per feature21,22,24,27e29,48, (4) sum scores
per joint group16,24,47,49, (5) mean score per feature12,30 or per
joint60, (6) scores on joint level (composite score or per
feature)12,20e24,34,35,38,40e44,47,48,51e53,60,61 and (7) presence or
absence of radiographic features per joint21,22,54,55,57,58, joint
group32,38,39,45, or on patient level52,56.

Discrimination

Reliability
Ten included articles provided data on the reliability of at least

two radiographic scoring methods, shown in Table III. The KL
scoring method was assessed in seven of these
studies12,16,17,20,21,23,24. Other assessed scoring methods were the
Kallman (n ¼ 4)12,17,20,23, OARSI (n ¼ 4)16,21,22,24, anatomical phases
(n ¼ 4)16,17,25,26, anatomical lesions (n ¼ 1)26, GUSS (n ¼ 1)25, global
score (n ¼ 1)17, and the semi-automated joint space width (JSW)
measurement (n ¼ 1).22

Eight studies provided cross-sectional data12,16,17,20e24. The ICCs
as well as kappa values were shown to be reliable for all assessed
total scores, and no differences between the scoring methods were
observed. The ICCs and kappa values for the individual radiographic
features depended on the scored feature; the lowest reliability was
reported for the scoring of cysts and the highest for the scoring of
erosions and osteophytes.12,20,21

In five of the studies readers performed the scoring indepen-
dently of another reader, providing results on the interreader reli-
ability12,16,17,21,24. The interreader ICCs and kappa values were
somewhat lower than the intrareader values, especially for the
Kallman method and for sclerosis as scored using the OARSI
atlas12,17,24. Whether readers were from one or different centers did
not seem to influence the reliability of the scoring methods.

Six studies provided data on the reliability of change of at least
two radiographic scoring methods12,16,17,24e26. The reliability of
change of KL, OARSI, Kallman, global, anatomical phases and GUSS
scores was reported to be good for all methods12,16,17,24e26. Bij-
sterbosch et al. compared the SDC of three scoring methods on



Table I
Radiographic scoring methods for hand osteoarthritis

Scoring method No. of
joints

DIP PIP IP1 MCP CMC1 STT Scored features Type of score Range of total
score

Anatomical phases14 26 þ þ þ þ e e Osteophytes, JSN, erosions, sclerosis Composite score 0e218.4
Anatomical lesions14 24 þ þ e þ e e Osteophytes, JSN, cysts Composite score Not specified
Burnett74 18 þ þ e e þ e Osteophytes, JSN, sclerosis Individual features 0e126
Eaton75 4 e e e e þ þ Osteophytes, JSN, erosions, cysts, sclerosis,

subluxation
Composite score Not specified

GUSS15 18 þ þ þ e e e Osteolytic areas, bone plate resorption, JSN Composite score 10e300
Kallman12 22 þ þ þ e þ þ Osteophytes, JSN, cysts, sclerosis, deformity,

cortical collapse
Individual features 0e208

Kellgren-Lawrence10 30 þ þ þ þ þ e Osteophytes, JSN, sclerosis, alignment Composite score 0e120
Kessler11 18 þ þ e e þ e Osteophytes, JSN, sclerosis Composite score 0e18
Lane76 22 þ þ þ e þ þ Osteophytes, JSN, erosions/cysts, sclerosis, deformity Individual features 0e182
OARSI13 20 þ þ þ e þ e Osteophytes, JSN, erosions/cysts, sclerosis, alignment Individual features 0e198

Abbreviations: CMC1 ¼ First carpometacarpal joint, DIP ¼ distal interphalangeal joint, IP1 ¼ First interphalangeal joint, MCP ¼ metacarpaphalangeal joint, No. ¼ number,
PIP ¼ proximal interphalangeal joint, STT ¼ scaphotrapezotrapezoidal joint.

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e1723 1713
patient level, showing a small difference in favor of the KL score,
followed by the anatomical phases and OARSI scores. Reported
SDCs were a little higher over a 6 year interval than over a 2 year
interval16. Haugen et al. assessed reliability of change in KL and
OARSI scores, showing a good reliability for the KL score and most
of the OARSI features. ICC and kappa values were somewhat lower
for change scores than for baseline KL and OARSI scores. Except for
change of sclerosis (OARSI), moderate to good reliability was re-
ported for the scoring of change in KL and OARSI scores24. Kallman
et al. evaluated agreement on progression in KL and Kallman scores
on joint group level, showing that agreement was more often
present in DIP joints than PIP joints and that agreement was lowest
on the progression of cysts.12

Sensitivity to change
Table IV shows the characteristics of the included studies

describing data on sensitivity to change of radiographic scoring
methods. Nine studies reported data on short-term follow-up (�3
years), most of them on patient level16,17,25e31. Two studies evalu-
ated change of summed KL, Kallman and anatomical phases scores,
of which one study also evaluated the global score16,17. Maheu et al.
reported SRMs over a 1 year interval of the global, KL, Kallman,
anatomical phases and OARSI scores; all below 0.50, indicating that
the responsiveness to change was small17. Bijsterbosch et al.
detected somewhat more progression over a 2 year interval when
scored following the KL or anatomical phases score as compared
with the OARSI atlas16. The anatomical phases score was evaluated
in two other studies25,26, one of these studies (a randomized
controlled trial (RCT)) also assessed change of GUSS. Progression
over a 1 year interval was detected by both scoring methods,
although no difference between treatment and placebo group was
observed.25

Five studies reported follow-up data of only one scoring meth-
od27e31. Botha-Scheepers et al. reported change of JSN and osteo-
phytes as scored following the OARSI atlas over a 2 year
interval27e29. Scoring of these features tended to be more sensitive
to change when scoring radiographs in chronological order as
compared with paired reading27. BucklandeWright et al. evaluated
stereoscopic measurement of individual OA features during a 1.5
year interval, reporting change of most features64. Olej�arov�a et al.
evaluated change of hand OA over a 2 year interval using the
Kallman scoring method, reporting no significant difference in total
score.31

In the three studies investigating long term follow-up data (>3
years), change in KL (n ¼ 2), OARSI (n ¼ 2), anatomical phases
(n ¼ 2) and anatomical lesions (n ¼ 1) score was evaluated12,14,16,24.
Studies with a longer follow-up duration detected higher
occurrence of progression of OA features as well as higher mean
radiographic change scores.16

Feasibility

Four studies reported data regarding feasibility of radiographic
scoring methods (Table V)11,16,17,22. The KL, anatomical phases and
Kallman scoring methods were assessed in two studies16,17. The
OARSI, Kessler and Lane scoring methods, as well as a non-validated
global score and semi-automated JSW measurement, were all
examined in only one study.11,16,17,22

The mean time to perform scoring ranged from 1.5 to 10e15 min
per hand radiograph. The KL, anatomical phases and Kessler scoring
methods seemed to be least time consuming while scoring ac-
cording Kallman, Lane and the OARSI atlas needed more time to
perform11,16,17. However, the time needed to perform the scoring
differed per study11,16,17. Bijsterbosch et al. showed that the per-
formance time increased in patients with higher levels of structural
abnormalities; 1 min increment in performance time was associ-
ated with 3.9 points in KL score (95% confidence interval (CI) 1.0,
6.8), 8.0 (5.3,10.7) points in OARSI score, and 21.1 (12.9, 29.2) points
in the anatomical phases scoring method.16

Validity

The 36 studies providing data regarding validity of radiographic
scoring methods are listed in Table VI. Analyses on individual joint
level were performed in 18 of these studies, and analyses on joint
group or patient level were performed in 13 and 14 studies,
respectively.

Thirteen studies focused on structural findings at physical ex-
amination in comparison to radiographic OA findings20,22,33e42.
Four studies presented correlation coefficients and kappa values,
reporting that nodes at physical examination were weakly to
moderately associated with radiographic hand OA34,35,37,38. The
lowest agreement was reported in a study on clinical Heberden
nodes and radiographic DIP osteophytes scored following the
Burnett scoring method, performed on joint level (k ¼ 0.36)35. The
highest correlation was reported in a study examining a clinical
score consisting of nodes and deformity and the radiographic KL
score, analyzed on joint group level (males r ¼ 0.47, females
r ¼ 0.66).38

Two studies reported the association between two radiographic
scoring methods and clinical nodes, both analyzed on a joint
level20,41. Addimanda et al., examining KL and Kallman scores, re-
ported the erosion and osteophyte features of the Kallman method
to be associated most strongly with nodes (OR 7.4 and 3.2



Table II
Overview of included studies (n ¼ 48)

First author, year
of publication

Source population, no. of patients (% women),
mean age (years)

Scoring methods Joints investigated Analysis of radiographic scores

Addimanda, 201220 Secondary care (50% erosive OA), 446 (93), 68 KL
Kallman

DIP, PIP, CMC1
DIP, PIP, CMC1

Score per joint, summed total
Score per joint per feature, summed per joint,
summed total

Bagge, 199133 General population, 217 (66), 82 KL DIP, PIP, IP, MCP, CMC1 Score per joint group (most affected joint)
Bijsterbosch, 201116 Familial polyarticular OA (GARP), 90 (78), 60 KL

OARSI
Anatomical phases

DIP, PIP, IP1, MCP, CMC1
DIP, PIP, IP1, CMC1
DIP, PIP, IP1, MCP

Summed per joint group, summed total
Summed per joint group, summed total
Summed per joint group, summed total

Botha-Scheepers,
200527

Familial polyarticular OA (GARP), 20 (90), median
age 62

OARSI DIP, PIP, IP1, MCP, CMC1,
STT

Summed total per feature

Botha-Scheepers,
200729

Familial polyarticular OA (GARP), 193 (80), 60 OARSI DIP, PIP, IP1, MCP, CMC1,
STT

Summed total per feature

Botha-Scheepers,
200928

Familial polyarticular OA (GARP), 172 (79), 61 OARSI DIP, PIP, IP1, CMC1 Summed total per feature

Buckland
eWright,199030

Unclear (radiographic OA patients), 32 (91), 62 Stereoscopic
measurement

DIP, PIP, MCP Mean score total per feature, mean score per joint
group per feature

Caspi, 200134 Secondary care (geriatric patients), 253 (68), 79 Modified OARSI DIP, PIP, IP1, MCP, CMC1 Score per joint, summed total
Ceceli, 201262 Secondary care, 60 (100), 59 Kallman Not specified Summed per hand
Cicuttini, 199835 General population (twin study), 660 (100), 56 Burnett

Kallman
DIP
PIP, CMC1

Score per joint
Score per joint

Dahaghin, 200443 General population (Rotterdam study), 3906
(58), 67

Modified KL DIP, PIP, MCP, CMC1, STT Score per joint, score per joint group, score per
patient (most affected joint)

Ding, 200744 Finnish dentists/teachers, 543 (100), range 45
e63

KL DIP, PIP, IP1, MCP Score per joint, no. of joints scored �2, summed
total

Dominick, 200545 Familial OA (Genetics of Generalized
Osteoarthritis (GOGO) study), 700 (80), 69

KL DIP, PIP, IP1, MCP, CMC1,
STT

Present/absent of score �2 per joint group,
summed total

Drape, 199632 Secondary care (mucoid cyst), 23 (61), 63 Osteophytes, JSN
(NVM)

DIP Present/absent per joint group per feature

El-Sherif, 200846 Secondary care, 40 (100), 57 KL DIP, PIP, IP1, MCP, CMC1 Score per patient (most affected joint)
Grainger, 200754 Secondary care, 15 (93), 59 Erosions (NVM) DIP, PIP Present/absent per joint
Hart, 199136 Primary/secondary care (non-joint related

problems), 541 (100), 54
KL DIP, PIP, CMC1 Score per joint group (most affected joint)

Hart, 199437 Primary care, 976 (100), age range 45e65 KL DIP, PIP, CMC1 Score per joint group (most affected joint)
Haugen, 201221 Secondary care (Oslo hand OA cohort),

106 (92), 69
KL
OARSI
Marginal erosions
(NVM)

DIP, PIP, IP1, MCP, CMC1
DIP, PIP, IP1, MCP, CMC1
DIP, PIP, IP1, MCP, CMC1

Score per joint, summed total
Score per joint per feature, summed total per
feature
Present/absent per joint

Haugen, 201324 Secondary care (Oslo hand OA cohort), 190 (91),
62 (longitudinal analysis: 99 (92), 61)

KL
OARSI

DIP, PIP, IP1, MCP, CMC1
DIP, PIP, IP1, MCP, CMC1

Score per joint, summed per joint group,
summed total
Score per joint per feature, summed total per
feature

Huetink, 201259 22 phantom joints, 22 human cadaver joints Automatic JSN
quantification

DIP, PIP, MCP Millimeter (mm) per joint

Iagnocco, 200556 Secondary care (inflam-matory OA), 110 (100),
67

Classical/erosive
OA (NVM)

DIP, PIP Present/absent per patient

Jones, 200147 Secondary care, 522 (67), 56 OARSI DIP, CMC1 Score per joint per feature, summed per joint
group

Jonsson, 201238 General population (AGES-Reykjavik study), 381
(58), 76

KL DIP, PIP, CMC1 Score per joint, present/absent of score �2 per
joint group, summed total

Kallman, 198912 General population (BLSA), 50 (0), 68 KL
Kallman

DIP, PIP, IP1, CMC1
DIP, PIP, IP1, CMC1, STT

Score per joint, score per joint group, mean score
total
Score per joint per feature, score per joint group
per feature, mean score total per feature

Keen, 200857 Secondary care, 37 (84), 57 OARSI DIP, PIP, MCP, CMC1 Present/absent per joint per feature
Kessler, 200011 Advanced hip/knee OA patients (Ulm OA study)

50, range 51e79
Kessler
Kallman
Lane

DIP, PIP, CMC1
DIP, PIP, CMC1
DIP, PIP, CMC1

No. of affected joints per joint group
Not specified
Not specified

Kortekaas, 201148 Secondary care, 55 (47), 61 OARSI DIP, PIP, IP1, CMC1 Score per joint per feature, summed total per
feature

Kwok, 201122 Familial polyarticular OA (GARP), 235 (83), 65,
and 471 controls

OARSI
Anatomical phases
Semi-automated
measured JSW

DIP, PIP, MCP
DIP, PIP
DIP, PIP, MCP

Score per joint per feature, summed total per
feature
Present/absent per joint
Score per joint, summed total

Lee, 201249 General population (KLoSHA), 378 (48), 75 KL DIP, PIP, IP1, MCP, CMC1 Summed per finger
Maheu, 200717 Secondary care, 105 (93), 61 KL

Kallman
Global score
Anatomical phases

DIP, PIP, MCP, CMC1
DIP, PIP, MCP, CMC1,STT
DIP, PIP, MCP, CMC1,
STT DIP, PIP, MCP

Summed total
Summed total
Summed total
Summed total

Mancarella, 201023 Secondary care, 35 (94), 66 KL
Kallman

DIP, PIP, MCP
DIP, PIP, MCP

Score per joint
Score per joint

Marshall, 200939 Primary care (hand pain), 592 (62), 64 KL DIP, PIP, IP1, MCP, CMC1,
STT

Present/absent of score �2 per joint group

Mathiessen, 201240 Secondary care (Oslo hand OA cohort), 127 (91),
69

OARSI DIP, PIP, IP1, MCP Score per joint per feature

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e17231714



Table II (continued)

First author, year
of publication

Source population, no. of patients (% women),
mean age (years)

Scoring methods Joints investigated Analysis of radiographic scores

Olej�arov�a, 200031 Secondary care, erosive OA: 28 (93), 68; non-
erosive OA: 24 (83), 65

Kallman DIP, PIP, IP1, MCP, CMC1 Summed total

Ozkan, 200750 Secondary care, 100 (87), 69 KL DIP, PIP, MCP, CMC1 Score per patient (most affected joint)
Rees, 201241 Secondary care (Genetics of Osteoarthritis and

Lifestyle (GOAL) study participants with �1
node), 1,939 (54), 68

KL
OARSI

DIP, PIP, IP1, CMC1
DIP, PIP, IP1, CMC1

Score per joint
Score per joint per feature

Saltzherr, 201361 Secondary care, 30 (70), median age 57 Eaton CMC1, STT Score per joint, score per joint per feature
Sonne-Holm, 200651 General population (Copenhagen city hearth

study), 3,355 (61),age>20
Modified KL CMC1 Score per joint, score per joint per feature

Stern, 200442 Primary and secondary care (Investigation of
Nodal Osteoarthritis to Detect an Association
with Loci encoding IL-1 (I-NODAL) study), 71
(80), 67

KL DIP, PIP, IP1, CMC1 Score per joint

Sunk, 201253 Post mortem IP joints, 40 (44), median age 66 KL
OARSI

DIP, PIP
DIP, PIP

Score per joint
Score per joint per feature

Verbruggen, 199614 Unclear (radiographic OA), 46 (96), 57 Anatomical phases
Anatomical lesions

DIP, PIP, MCP
DIP, PIP, MCP

Summed total
Summed total

Verbruggen, 200226 Unclear (radiographic OA, two RCT's), 222 (92),
56

Anatomical phases
Anatomical lesions

DIP, PIP, MCP
DIP, PIP, MCP

Summed total
Summed total

Verbruggen, 201225 Secondary care (RCT), 60 (85), 61 Anatomical phases
GUSS

DIP, PIP
DIP, PIP

No. of joints in each phase per patient
Summed total

Van ‘t Klooster,
200860

Familial polyarticular OA (GARP), 40 (33), 60 OARSI
Automatic JSW
quantification

DIP, PIP, MCP
DIP, PIP, MCP

Score per joint
Mean score per joint

Vlychou, 200958 Secondary care (OA patients), 22 (91), 63 Osteophytes,
erosion (NVM)

DIP, PIP, IP1, MCP, CMC1 Present/absent per joint per feature

Wittoek, 201155 Secondary care, erosive OA: 9 (67), median 61;
non-erosive OA: 5 (100), median 63

Osteophytes,
erosions (NVM)

DIP, PIP Present/absent per joint per feature

Zhang, 200252 General population (Framingham hand OA
study), 1,032(64), age�71

Modified KL DIP, PIP, IP1, MCP, CMC1 Score per joint, present/absent of score �2 per
patient

Abbreviations: AGES ¼ Age, Gene/Environment Susceptibility, BLSA ¼ Baltimore Longitudinal Study of AgingGARP ¼ Genetics osteoarthritis and Progression,
KLoSHA ¼ Korean Longitudinal Study on Health and Aging, NVM ¼ non-validated method, OA ¼ osteoarthritis, .

Table III
Studies providing data on reliability of scoring methods (n ¼ 10)

First author No. of readers, centers Intrareader reliability* Interreader reliability*

Cross-sectional studies
Addimanda20 2 (consensus), 1 KL: ICC 0.994

Kallman: ICC 0.987, k range per feature 0.42e0.81
N/A

Bijsterbosch16 3 (independent), 3 KL: ICC range per reader 0.90e0.96
OARSI: ICC range per reader 0.77e0.97
Anatomical phases: ICC range per reader 0.88e0.97

KL: ICC range per two readers 0.84e0.91
OARSI: ICC range per two readers 0.80e0.95
Anatomical phases: ICC range per two readers
0.81e0.95

Haugen21 2 (independent), 2 KL: ICC 0.97, k 0.86 (one reader)
OARSI (including marginal erosions):
ICC range per feature 0.70e0.97, k range per feature 0.75e0.88
(one reader)

KL: ICC 0.96, k 0.79
OARSI (including marginal erosions):
ICC range per feature 0.56e0.95, k range per feature
0.62e0.81

Haugen24 2 (independent), 2 KL: ICC 0.97, k 0.82 (one reader)
OARSI: ICC range per feature 0.62e0.96, k range per feature 0.64e0.81
(one reader)

KL: ICC 0.95, k 0.70
OARSI: ICC range per feature �0.07e0.94,
k range per feature 0.00.-0.77

Kallman12 4 independent, 2 KL mean score: ICC 0.80, range per joint group 0.68e0.87
Kallman mean score: ICC per feature range 0.74e0.84, per feature per
joint group range 0.62e0.93

KL mean score: ICC 0.74, range per joint group
0.74e0.81
Kallman mean score: ICC per feature range 0.29e0.71,
per feature per joint group range 0.33e0.82

Kwok22 2 (consensus), 1 OARSI (JSN): ICC 0.92
Semi-automated JSW: ICC 0.99, mean difference 0.017 mm (standard
deviation (SD) 0.04), smallest detectable difference (SDD) 0.055 mm

N/A

Maheu17 2 (independent), 2 KL: ICC range per reader 0.988e0.991
Kallman: ICC range per reader 0.962e0.999
Global: ICC range per reader 0.922e0.961
Anatomical phases: ICC range per reader 0.999e0.999

KL: ICC 0.951
Kallman: ICC 0.706
Global: ICC 0.859
Anatomical phases: ICC 0.996

Mancarella23 2, not specified KL: ICC score per joint 0.99
Kallman: ICC score per joint 0.99

Longitudinal studies
Bijsterbosch16 3 (independent), 3

Mean follow-up 2 years
Mean follow-up 6 years

KL: SDC range per reader 2.1e7.1
OARSI: SDC range per reader 1.2e10.2
Anatomical phases: SDC range per reader 1.4e7.8
KL: SDC range per reader 3.7e8.1
OARSI: SDC range per reader 3.0e11.1
Anatomical phases: SDC range per reader 3.5e9.9

KL: SDC 2.9
OARSI: SDC 4.1
Anatomical phases: SDC 2.7
KL: SDC 3.8
OARSI: SDC 4.6
Anatomical phases: SDC 4.0

(continued on next page)

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e1723 1715



Table III (continued)

First author No. of readers, centers Intrareader reliability* Interreader reliability*

Haugen24 2 (independent), 2
Mean follow-up 7 years

KL: ICC 0.93, k 0.83 (one reader)
OARSI: ICC range per feature �0.02e0.96,
k range per feature 0.00e0.90 (one reader)

KL: ICC 0.83, k 0.53
OARSI: ICC range per feature �0.03e0.90,
k range per feature �0.03e0.71

Kallman12 4 (independent), 2
Mean follow-up 23
years

N/A KL: scattered agreement
Deformity/collapse: agreement
Cysts: disagreement
Osteophytes/JSN/sclerosis: scattered agreement

Maheu17 2 (independent), 2
Mean follow-up 1 year

KL: ICC range per reader 0.990e0.998
Kallman: ICC range per reader 0.986e0.959
Global: ICC range per reader 0.939e0.956
Anatomical phases: ICC range per reader 0.941e0.988

KL: ICC 0.998
Kallman: ICC 0.995
Global: ICC 0.999
Anatomical phases: ICC 0.998

Verbruggen26 2 (independent), 1
Mean follow-up 3 years

Anatomical phases: agreement for two RCTs 84e93%, k 0.6e0.8
Anatomical lesions: correlation for two RCTs r 0.7e0.9, R2 44e87%

Anatomical phases: agreement for two RCTs 81e85%,
k 0.6e0.7
Anatomical lesions: correlation for two RCTs r
0.7e0.8, R2 55e66%

Verbruggen25 2 (independent), 1
Mean follow-up 1 year

Anatomical phases: 96% agreement, k 0.95
GUSS: ICC 0.97

Anatomical phases: 94% agreement, k 0.92
GUSS: ICC 0.86, SDC 18

Abbreviations: k ¼ kappa, , N/A ¼ not applicable, R2 ¼ explained variance.
* Unless stated otherwise ICCs are for summed total scores on patient level, k's on joint level.

Table IV
Studies providing data on sensitivity to change of radiographic scoring methods in hand osteoarthritis (n ¼ 11)

First author Mean follow-up
(years)

Definition of progression Sequence known/
unknown

Results relevant for evaluation of sensitivity to change

Short-term
Bijsterbosch16 2 Change > SDC Known Percentage progression (range for three readers):

- KL: 19e56%
- OARSI: 7e38%
- Anatomical phases: 13e52%

Botha-
Scheepers27

2 �1 score Known/
unknown

Progression of JSN/osteophytes:
- chronological reading: 1/15% (SRM 0.38/0.41)
- paired reading: 5/15% (SRM 0.00/0.39)

Botha-
Scheepers28

2 �1 score Unknown JSN: 19% progression, mean change 0.3, SRM 0.34
Osteophytes: 22% progression, mean change 0.4, SRM
0.35

Botha-
Scheepers29

2 �1 score Unknown JSN: 24% progression (�2/�3/�4 score: 10/4/3%)
Osteophytes: 22% progression (�2/�3/�4 score: 10/4/
3%)

Buckland-
Wright30

1.5 Change > variations in precision Not specified JSW: 62% narrowing (P < 0.02)
Subchondral sclerosis: 60% increase, 34% decrease
Osteophytes: increase in size and no. (P < 0.005)
Juxta-articular radiolucencies: increase in size
(P < 0.002), not in no.

Maheu17 1 Change in
summed score

Unknown SRM for two readers:
- KL: 0.17/0.24
- Kallman: 0.26/0.29
- Global: 0.17/0.27
- Anatomical phases: 0.18/0.27

Olej�arov�a31 2 Change in
summed score

Unknown Erosive OA: change 5.0, P > 0.05
Non-erosive OA: change 4.3, P > 0.05

Verbruggen26 3 Change in anatomical phases,
Change in anatomical lesions

Known Anatomical lesions showed different progression
between trial
arms, anatomical phases did not.

Verbruggen25 1 Change in anato-mical N/S/J phase to E
phase,
Change in summed score

Unknown No. (%) joints with progression to E phase:
- Total group: 24 (2.8%) of 848 N/S/J joints
- Placebo treated: 15 (3.6%) of 429 N/S/J joints
- Adalimumab treated: 9 (2.1%) of 419 N/S/J joints
Mean difference GUSS (baseline palpable swelling yes/
no):
- Placebo: �5/3
- Adalimumab: 4/1

Long-term
Bijsterbosch16 6 Change > SDC Known Percentage progression (range for three readers):

- KL: 51e80%
- OARSI: 33e74%
- Anatomical phases: 27e66%

Haugen24 7.3 Change in score Known Progression (percentage of joints):
- KL: 29%
- OARSI: osteophytes 19%, JSN 13%, erosions 9%, mala-
lignment 4%, cysts 2%, sclerosis 1%

Verbruggen14 4.6 Change in anatomical phases,
Change in anatomical lesions

Known Progression of anatomical lesions more frequent in PIP/
DIP than MCP. Progression of anatomical phases in 43%.
Progression according anatomical phases and
anatomical lesions yielded comparable results.

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e17231716



Table V
Studies providing data on feasibility of radiographic scoring methods in hand
osteoarthritis (n ¼ 4)

First author No. of
radiographs

Mean (SD) time to perform scoring

Bijsterbosch16 3 KL: 4.3 (2.5) min
OARSI: 9.3 (6.0) min
Anatomical phases: 2.8 (1.5) min

Kessler11 1 Kessler: 5 min per hand
Kallman: 10e15 min per hand
Lane: 10e15 min per hand

Kwok22 1 Semi-automated JSW measurement: 5.1 (2.8)
min

Maheu17 1 KL: 1.9 (0.6) min
Kallman: 3.5 (0.7) min
Global score: 1.5 (0.5) min
Anatomical phases: 1.6 (0.5) min

Abbreviations: min ¼ minutes, no. ¼ number.

Table VI
Studies providing data on validity of scoring methods (n ¼ 37)

First author Validation method

Clinical: structural findings at physical examination
Addimanda20 Heberden/Bouchard nodes (yes/no)

Bagge33 Nodes/periarticular enlarge-ment, instability, squaring
(yes/no �1 feature per joint)

Caspi34 Nodes, malalignment DIP/PIP (summed)

Cicuttini35 Heberden nodes (yes/no)
Hart36 Nodes (yes/no)

Hart37 Nodes IP (graded 0e4), squaring CMC1 (grade 0e1)

Jonsson38 Nodes, deformity
(graded 0e3, summed)

Kwok22 Nodes (yes/no)

Marshall39 Nodes, deformity, enlargement (yes/no)

Mathiessen40 Nodes (yes/no)
Rees41 Nodes (yes/no)

Stern42 Nodes (yes/no)

Clinical: symptoms, function
Bagge33 Pain/stiffness (interview, yes/no)

Ceceli62 Pain (visual analog scale(VAS)), disability (Disabilities of
the Arm Shoulder and Hand (DASH) questionnaire),
dexterity (Purdue pegboard test), grip/pinch strength

Dahaghin43 Pain (interview, yes/no)/disability (Health Assessment
Questionnaire (HAQ))

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e1723 1717
respectively)20. Rees et al. examined the association between KL
and OARSI scores and clinical nodes, reporting ORs only for the KL
method (range per joint 2.3e21.2). Regarding the OARSI atlas, JSN
was mentioned to be more strongly associated with clinical nodes
than osteophytes.41

Seventeen studies assessed clinical symptoms and hand func-
tion in comparison to radiographic scoring methods (KL: n ¼ 14,
OARSI: n ¼ 3, Kallman: n ¼ 1, JWS/JSN: n ¼ 1)22,24,33,36,37,39,43e52,62.
All studies reported significant associations between radiographic
OA features and pain and disability, of which four showed a dose-
dependent association between KL and OARSI scores and
pain24,43,44,48. Of the nine studies assessing grip or pinch strength,
only two did not find an association with radiographic OA (1x KL,1x
JSW/JSN, analyzed on patient level).22,50

Only one study assessed longitudinal data, showing incident or
progressive KL or OARSI scores to be associated with incident pain
Results relevant for evaluation of validity

OR (95% CI) for nodes on joint level, adjusted for disease duration, body mass index
(BMI):
- KL: 2.20 (2.09, 2.31)
- Kallman: 1.17 (1.62, 1.72)
- Kallman JSN: 2.57 (2.40, 2.75)
- Kallman osteophytes: 3.19 (2.97, 3.42)
- Kallman central erosions: 7.4 (6.0, 10.1)
Correlated with KL score in all joint groups (correlation coefficient not provided),
test for linear trend: P < 0.01.
Clinical features also present in KL 0 joint groups.
Correlation with OARSI:
- summed total: r 0.4 (P 0.001)
- DIP/PIP: range per joint r 0.18e0.52 (P 0.004e0.0001)
k with DIP osteophytes (Burnett): 0.36 (95% CI 0.33, 0.39)
Sensitivity for KL �2: range per joint group 19e49%
Specificity for KL �2: range per joint group 87e98%
Prevalence node �2: KL0: 3%, KL1: 19%, KL2: 48%, KL3: 74%, KL4: 82%
Prevalence squaring: KL0: 5%, KL1: 11%, KL2: 25%, KL3: 41%, KL4: 70% (correlation
coefficient not specified)
Correlation summed score with summed total KL: males r 0.47, females r 0.66
Prevalence KL � 2 (DIP 67%, PIP 32%, CMC1 20%) higher as compared to clinical
grade �2 (DIP 54%, PIP 19%, CMC1 10%)
b (95% CI) for nodes on joint level, adjusted for age, sex, BMI, family effect, mean
phalanx width:
- JSW: �0.37 (�0.40, �0.34)
- JSN: 0.48 (0.42, 0.55)
OR (95% CI) of presence of �1 feature for:
- KL � 2 in CMC1: 2.2 (1.5, 3.3)
- KL � 2 in any thumb joint: 3.1 (2.1, 4.5)
Osteophytes (OARSI) in 30% of joints, nodes in 37% of joints
KL � 2 associated with any node on patient level: OR range per joint 2.26e21.23
(adjusted for age, sex, BMI, hand dominance, trauma, occupation, sports)
JSN/osteophytes (OARSI) also associated with nodes (P < 0.001); ORs of JSN
greater than ORs of osteophytes in all joints except for IP1/CMC1
Sensitivity for KL � 2: range per joint group 42e100%
Specificity for KL � 2: range per joint group 17e94%

Correlated with KL score in all joint groups (correlation coefficient not provided),
test for linear trend: P < 0.01.
Correlation with summed Kallman score right/left hand:
- Pain: r 0.17/0.18 (P > 0.05)
- Disability: r 0.29/0.30 (P < 0.05)
- Dexterity: r �0.26/�0.30 (P < 0.05)
- Grip strength: r �0.37/�0.40 (P < 0.05)
- Pinch strength: r range per test �0.31 to �0.25/�0.35 to �0.27 (P < 0.05)
OR (95% CI) for KL � 2/�3/4 on patient level, adjusted for age, sex:
- pain: 1.9 (1.5, 2.4)/1.8 (1.3, 2.5)/3.6 (2.2, 5.8)
- disability: 1.5 (1.1, 2.1)/1.6 (1.1, 2.5)/1.6 (0.9, 2.9)

Pain associated with KL � 2 in PIP/CMC1/STT, disability with KL � 2 in MCP
Adjusted OR (95% CI) for KL � 2 in all joint groups: pain 2.7 (1.4, 5.2), disability 2.7
(1.3, 6.0)

(continued on next page)



Table VI (continued )

First author Validation method Results relevant for evaluation of validity

Ding44 Pain (questionnaire, yes/no per joint, summed) Correlation with summed total KL: r 0.26 (P 0.0005)
Correlation with no. KL � 2 joints: r 0.28 (P 0.0005)
prevalence ratio (PR) (95% CI) for pain on joint level, adjusted for age, occupation:

- KL 2: 1.70 (1.44, 2.01)
- KL � 3: 5.17 (4.34, 6.16)
Adjusted PR (95% CI) for mild/moderate pain on joint level:

- KL 2: 1.93 (1.54, 2.41)/2.21 (1.58, 3.10)
- KL � 3: 4.92 (3.77, 6.43)/11.73 (8.95, 15.38)

Dominick45 Grip/pinch strength b (P-value) for grip/pinch strength, adjusted for age, sex, pain, chondro-calcinosis,
hand hypermobility:
- Summed total KL: �0.67 (<0.001)/�0.16 (<0.001)
- KL � 2 PIP: �6.67 (0.003)/�1.17 (0.070)
- KL � 2 MCP: �3.32 (0.114)/�1.78 (0.003)
- KL � 2 CMC: �9.06 (<0.001)/�1.03 (0.049)
- KL � 2 per finger: range �1.81 to �11.08 (P < 0.05)

El-Sherif46 AUSCAN, morning stiffness (minutes), grip strength,
Ritchie index

AUSCAN pain/function higher in KL4 than KL2 (P < 0.05)
Correlation with KL score:

- AUSCAN pain: r 0.459 (P 0.003), function: r 0.394 (P 0.012)
- Grip strength right hand: r �0.322 (P 0.043)
Other measures not significantly correlated with KL

Hart36 Tenderness, pain on movement (physical examination,
yes/no)

Comparison tenderness/pain on movement with KL � 2:
- sensitivity: range per joint group 7e26%/1e22%
- specificity: range per joint group 92e99%/96e99%

Hart37 Pain, stiffness (interview, yes/no) Prevalence symptoms in patients with KL < 2: 15%, KL2: 49%, KL3-4: 81%;
test for linear trend: P < 0.01

Haugen24 Tenderness on palpation (yes/no), grip strength,
AUSCAN

Cross-sectional OR (95% CI) for tenderness on joint level, adjusted for age, sex:
- KL score 1/2/3/4: 1.4 (1.2, 1.7)/3.0 (2.4, 3.7)/6.8 (4.5, 10)/5.3 (3.3, 8.6)
- OARSI osteophytes score 1/2/3: 2.8 (2.3, 3.4)/4.3 (3.0, 6.3)/4.5 (2.9, 7.0)
- OARSI JSN score 1/2/3: 0.9 (0.7, 1.2)/1.9 (1.4, 2.5)/2.5 (1.7, 3.7)
- OARSI erosions: 3.3 (2.3, 4.9), malalignment: 2.8 (2.0, 3.9), cysts: 2.2 (1.4,3.3),
sclerosis: 2.6 (1.1, 6.0)

AUSCAN pain associated with summed KL and OARSI osteophytes/JSN. AUSCAN
function associated with summed KL and OARSI osteophytes, JSN, erosions, cysts.
Grip strength associated with summed KL and all OARSI features except for
sclerosis.
Summed KL per joint group only associated with grip strength (CMC1 strongest)
Adjusted OR (95% CI) of progressive/incident scores for incident tenderness:
- KL score 1/2/3/4: 1.2 (0.7, 2.0)/1.5 (0.9, 2.4)/5.7 (3.0, 11)/11 (4.0, 33)
- OARSI osteophytes: 3.0 (2.0, 4.4), JSN: 2.8 (1.7, 4.7), erosions: 8.4 (4.7, 15),
malalignment: 3.8 (1.9, 7.4), cysts: 2.2 (0.9, 5.0), sclerosis: 2.4 (0.8, 8.0)

Increasing summed KL and OARSI JSN/malalignment associated with increased
AUSCAN function. More malalignment associated with less grip strength
Change summed KL per joint group not associated with AUSCAN/grip strength

Jones47 AUSCAN, grip strength Association with summed OARSI per joint group, adjusted for age/sex/other joints/
Heberden nodes:
- AUSCAN pain: PIP b 0.17, CMC1 b 0.14 (P < 0.05)
- AUSCAN function: PIP b 0.15, CMC1 b 0.19 (P < 0.05)
- grip strength: PIP b �0.12, CMC1 b �0.09 (P < 0.05)

Kortekaas48 AUSCAN, pain (VAS), Doyle index of hands OR (95% CI) for pain on palpation on joint level, adjusted for age, sex, BMI:
- osteophytes score 1/2/3: 2.2 (1.7, 2.9)/3.9 (2.6, 5.9)/4.8 (2.7, 8.4)
- JSN score 1/2/3: 2.0 (1.4, 2.8)/5.3 (3.1, 9.1)/6.4 (2.7, 14.8)

Summed osteophytes/JSN not associated with AUSCAN pain, VAS, Doyle.
Kwok22 AUSCAN, pain on palpation (yes/no), grip strength,

mobility
b (95% CI) for JSW/JSN on joint level, adjusted for age, sex, BMI, family effect, mean
phalanx width:
- self-reported pain: �0.21 (�0.27, �0.16)/0.39 (0.30, 0.48)
- pain on palpation: �0.25 (�0.29, �0.21)/0.37 (0.29, 0.44)
No. joints with self-reported pain/pain on palpation, AUSCAN pain/function and
mobility associated with summed JSW/JSN. Grip strength not associated

Lee49 Grip/pinch strength, disability (DASH questionnaire) Associations with summed KL, adjusted for age/sex (P < 0.05):
- grip strength: thumb b �1.05, third finger b �2.17
- pinch strength: thumb b �0.28, second finger b �0.26
-disability: thumb b 1.53, second finger b 0.63, third finger b 3.97

Marshall39 AUSCAN, pain during activity/pain in past month
(questionnaire, yes/no), grip/pinch strength, grind test,
Finkelstein's test

OR (95% CI) for KL � 2 in CMC1/any thumb joint:
- Pain during activity: 2.1 (1.5, 2.9)/2.2 (1.6, 3.2)
- Pain in past month: 1.5 (1.0, 2.1)/1.4 (1.0, 2.0)
- Grind test: 1.8 (1.1, 2.9)/1.7 (1.0, 2.9), Finkelstein's test not associated

Ozkan50 Grip/pinch strength, Dreiser's functional index,
disability (HAQ)

Disability KL score <2/2/3-4: 2.40/2.10/6.45 (KL3-4 vs KL < 2/2 P < 0.05)
Dreiser's index KL score <2/2/3-4: 2.73/2.10/9.25 (KL3-4 vs KL < 2/2 P < 0.05)
Grip/pinch strength not different between KL scores

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e17231718



Table VI (continued )

First author Validation method Results relevant for evaluation of validity

Sonne-Holm51 Pain CMC1 (interview, yes/no) OR (95% CI) for pain, adjusted for age, sex, BMI:
- KL: 1.48 (1.33, 1.65)
- Sclerosis/cyst: 1.48 (1.23, 1.77)/1.23 (1.03, 1.47)

JSW and osteophytes not associated.
Zhang52 Functional limitations (questionnaire), grip strength Patients with KL � 2 and joint pain/aching/stiffness had more functional

limitations and lower grip strength; age adjusted difference (95% CI) men 3.1 kg
(1.8, 4.4), women 1.9 kg (1.4, 2.4)

Histological
Sunk53,69 Modified Mankin score (range 0e14; >5 ¼ OA) Correlation with KL score (DIP/PIP): r 0.87/0.79 (P < 0.0001)

Correlation with OARSI JSN: r 0.77/0.76, osteophytes: r 0.89/0.69 (P < 0.0001)
Sensitivity KL � 2 for Mankin >5 (DIP/PIP): 84.6/54.2%, specificity: 100/100%

MRI
Drape32 Pedicled cysts DIP (yes/no) 19 pedicled cysts: 16 associated with osteophytes/JSN on CR, three no

osteophytes/JSN on CR
Grainger54 Erosions (central/marginal, yes/no) 37 MRI erosions: 24% also on CR (44% of central, 5% of marginal erosions)

All CR erosions also on MRI
Haugen21 Oslo hand OA score

(graded per feature)
Agreement with osteophytes k 0.41, JSN k 0.50, central erosions k 0.75, central/
marginal erosions k 0.43, cysts k 0.11, malalignment k 0.50

Wittoek55 Erosions, osteophytes (yes/no) Prevalence erosions: MRI PIP 29%, DIP 68%, CR PIP 11%, DIP 38%
PIP osteophytes (erosive/non-erosive) hand OA MRI 25/50%, CR 42/40%
DIP osteophytes: MRI and CR >80%

CT
Saltzherr61 JSN, osteophytes, subchon-dral sclerosis, cyst, erosion,

subluxation (OA defined on no. of features)
Prevalence of individual features and OA higher according to CT than CR

US
Iagnocco56 Erosions (yes/no) US erosions in 16 (72.7%) of 22 CR erosive hand OA patients.

No US erosions in CR classical hand OA patients (n ¼ 88).
Keen57 JSN, osteophytes (yes/no) Osteophytes: k 0.54 (77.8% agreement)

JSN: k 0.436 (74.6% agreement)
Kortekaas48 Osteophytes (yes/no) US osteophytes 69%, OARSI osteophytes 46%
Mancarella23 Cartilage thickness (mm) Negatively correlated with KL and Kallman score (P < 0.0001)
Mathiessen40 Osteophytes (yes/no) OARSI osteophytes in 30% of joints, US osteophytes in 53% of joints

CR and US: 57.3% exact agreement, 88.3% close agreement
Vlychou58 Central erosions, osteophytes (yes/no) CR detected less erosions/osteophytes (17/47%) than US (35/55%), P < 0.05

Difference most apparent in DIP and PIP
Wittoek55 Erosions, osteophytes (yes/no) CR detected less erosions (PIP 11%, DIP 38%) than US (21, 52%) in erosive and non-

erosive hand OA
CR detected less PIP osteophytes (41%) than US (54%).
CR and US both detected >80% DIP osteophytes

Digital photography
Jones47 Heberden nodes (yes/no) Correlation with OARSI score �1 in DIP joints: r 0.74 (P < 0.001)
Jonsson38 Tissue enlargement/deformity (graded 0e3 per joint,

summed)
Prevalence OA higher according to KL � 2 (DIP 67%, PIP 32%, CMC1 20%) as
compared to digital photograph �2 (DIP 33%, PIP 20%, CMC1 3%)
Correlation summed score with summed total KL: males r 0.35, females r 0.53

Stern42 Hard tissue enlargement (yes/no) Sensitivity for KL � 2: range per joint 17e74%
Specificity for KL � 2: range per joint 67e92%

Other measures of JSW
Huetink59 True JSW by micrometer Compared to automatic JSN quantification:

Mean difference (SD): phantom joints: 0.052 (0.014) mm, cadaver joints:
0.210 (0.115) mm
SDD: phantom joints 0.028 mm, cadaver joints: 0.226 mm

van't Klooster60 Automatic JSW quantification (mm) Association with OARSI JSN: R2 0.54, P < 0.01

Abbreviations: kg ¼ kilogram, r ¼ correlation coefficient.

A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e1723 1719
on joint level and with change in Australian/Canadian Hand Oste-
oarthritis Index (AUSCAN) pain/function and grip strength.24

One study examined the association between the KL and OARSI
scoring methods and histological findings on joint group level,
showing a good correlation (r � 0.7) as well as a high sensitivity and
specificity.53

Four studies assessed individual features of hand OA by both
radiography and MRI21,32,54,55. The agreement between the two
methods was lowest for the presence of cysts and highest for
central erosions21. Three of the studies showed that MRI detected
more osteophytes, cysts and erosions as compared to
radiography.32,54,55

One study assessed individual features of CMC1 and STT OA by
both radiography and CT, reporting the latter to detect more JSN,
osteophytes, subchondral sclerosis, cysts, erosions and
subluxation.61
Seven studies used both US and radiography to assess hand OA
signs23,40,48,55e58. Six of the studies examined individual radio-
graphic features and reported US to detect more osteophytes and
erosions than radiography. A study on KL and Kallman scores re-
ported a negative correlation between radiographic JSN and US-
detected cartilage thickness on joint level.23

Three studies examined hand OA using digital photography and
radiography38,42,47. Two studies, performed on joint group level,
reported a good correlation between OARSI scores and Heberden
nodes on digital photography (r ¼ 0.74), and a weak to moderate
correlation between summed KL scores and summed digital
photograph score (comprising enlargement and deformity) on
digital photography (males r ¼ 0.35, females r ¼ 0.53).38,47

Finally, two studies examined quantitative measures of JSW,
both on individual joint level59,60. Van't Klooster et al. showed that
automatic JSW quantification was associated with JSN scored



A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e17231720
according to the OARSI atlas60. Huetink et al. reported that auto-
matic JSW quantification has a high accuracy in measuring the true
JSW (assessed by micrometer).59

Discussion

This review aimed at evaluating the radiographic scoring
methods used in hand OA research and to assess their metric
properties. We noticed that a wide variety of scoring methods has
been used in studies evaluating radiographic hand OA. Further-
more, the joints that were examined and the analysis of the ob-
tained scores differed extensively across studies. Evaluation of
metric properties of the evaluated scoring methods regarding
reliability, sensitivity to change, feasibility and validity did not
reveal major differences.

Both intra- and interreader reliability of all evaluated radio-
graphic scoring methods were good for summed scores and global
scores, for both cross-sectional and longitudinal radiographic
scoring. When grading individual radiographic features, the highest
reliability was reported for the scoring of erosions and osteophytes
and the lowest for the scoring of cysts.

When evaluating sensitivity to change, only one study evaluated
this in different groups of patients (trial arms) using different
scoring methods. Although such comparative studies may provide
the best insights in sensitivity to change, the included observational
follow-up studies showed the ability to detect change in structural
damage over time with CR. Change over time was observed even in
short term follow-up studies (<3 years). Reported SRMs were
similar for all evaluated scoring methods.

The feasibility of scoring methods has been described in a
limited number of studies. The performance time of the scoring
differed not only across the evaluated scoring method but also
across studies, and was shown to increase with the amount of
structural damage.

A large number of studies investigated the validity of radio-
graphic OA findings in comparison with clinical findings at physical
examination (such as nodes and deformities) and symptoms and
function; there was large variation between these studies. This
could be due to the various analyses of radiographic and clinical
findings, e.g., joint level vs patient level, and individual features vs
summed scores. Furthermore, studies were difficult to compare
because of the use of different effect measures, such as odds ratios
(ORs), correlation coefficients, sensitivity and specificity. In general
we can say that there was moderate agreement between radio-
graphic features and structural findings at physical examination.
The association of radiographic findings with hand function and
symptoms was reported to be stronger than the association with
findings at physical examination. All evaluated radiographic scores
were associated with grip strength and pain, the relation with pain
was observed on joint level as well as on patient level, and was
shown to be dependent on the radiographic severity. No differences
between the evaluated radiographic scoring methods were
observed. Only few studies assessed longitudinal associations
between radiography and pain or function, requiring further
validation.

In comparison with other imaging methods, radiography
appeared to detect fewer structural damage than MRI, CT and US,
and more structural damage than digital photography. However,
the findings on MRI, CT and digital photography require further
confirmation because of limited evidence. Agreement between
radiography and other imaging methods was assessed most often
on joint level and differed per feature.

Although no major differences regarding the metric properties
of the evaluated radiographic scoring methods were observed in
this review, the examined joints and analysis of the obtained scores
were shown to differ extensively across studies. All kinds of pre-
sentation of radiographic outcome measures were used, such as
scores per joint, summed scores, presence/absence of radiographic
OA features, or the highest scored joint. Summed scores were used
most frequently for evaluation of the reliability of radiographic
scoring methods and change of structural damage over time,
analyzed on patient level. When evaluating the validity of scoring
methods, analyses on individual joint level or on joint group level
were performed most often.

The various examined joints within hand OA research has been
described before in a review by Marshall et al. In addition, they
evaluated the use of definitions of hand OA, reporting some
agreement in the definition of individual joint OA but a wide
variation in defining overall hand OA65. Kerkhof et al. showed that
the use of varying definitions of radiographic OA within the same
study leads to different results66. Therefore, as stated before by
Haugen et al., standardization of the evaluation and definition of
radiographic hand OA with respect to scoring methods, examined
joints and required number of affected joints could reduce the
variation across studies.67

Based on this review, it is not possible to decide on what
radiographic scoring method should be recommended in hand
OA research. Although no major differences regarding metric
properties of the scoring methods were observed, the amount of
supporting evidence differed for the evaluated methods, which
may provide an argument for recommendation of specific scoring
methods. Most evidence across all evaluated domains is available
for the KL and OARSI scoring methods. Although global scoring
methods may be more reliable than the scoring of individual
radiographic features, individual features may be more suitable
for evaluation of specific study objectives. Therefore, the OARSI
scoring method may be recommended for evaluation of indi-
vidual radiographic features in addition to use of the KL scoring
method for global radiographic assessment. The OARSI Task Force
recommendations for the design and conduct of clinical trials in
hand OA already stated that the use of either aggregate radio-
graphic scores or grading of individual features depends on the
aim of study9. However, consensus should be reached on a more
specific definition; when should a global or individual feature
score be used and what specific scoring method should be rec-
ommended. Furthermore, consensus on the evaluated joints,
presentation of the radiographic outcome measures and the
definition of hand OA will help to enhance the comparability of
studies in hand OA.

A limitation of this study is that the methodological quality of
the included studies was not assessed, due to the heterogeneity
across studies regarding their purpose. The heterogeneity
regarding examined joints and analyses of obtained radiographic
scores did not enable performance of a meta-analysis. Furthermore,
publication bias was not addressed.

Although we aimed to provide a comprehensive overview of
available literature, the formulated inclusion and exclusion criteria
resulted in a specific selection of studies.

Consequently, some radiographic scoring methods were not
included in this review, being the Eaton-Littler classification system
and the recently developed interphalangeal OA radiographic
simplified (iOARS) score. These methods have not been evaluated
for reliability together with another method.68,69

Since sensitivity to change was evaluated in follow-up studies
assessing hand OA by at least two radiographic scoring methods in
case of long-term follow-up studies (>3 year), a number of studies
or abstracts evaluating change in KL and OARSI scores could not be
included.3,70e72

In the evaluation of the feasibility of the available radiographic
scoring methods in hand OA, we did not focus on the importance of



A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e1723 1721
radiographic techniques. Dela Rosa et al. evaluated the reliability of
scoring OA of the CMC1s according to the Eaton method when
using different X-ray views, showing that a combination of the
posterior-anterior, lateral and Bett's view showed a higher reli-
ability than using only one or two views73. Standardization of
radiographic techniques might further enhance comparability of
studies in hand OA.

In conclusion, this systematic review provides an overview of
the radiographic scoring methods used in the assessment of
structural damage in hand OA. We showed that several scoring
methods are available, evaluation of their metric properties
regarding reliability, sensitivity to change, feasibility and validity
did not reveal major differences. The examined joints and analysis
of the obtained radiographic scores differed extensively across all
studies. To enhance comparability across studies in hand OA,
consensus has to be reached on a preferred scoring method, as well
as on the examined joints and the used outcome measure.

Contributions

Authors made substantial contributions to the following: (1a)
conception and design of the study: AWV, PB, DMH, MK; (1b)
acquisition of data: AWV, JWS, MK; (1c) analysis and interpretation
of data: AWV, PB, IKH, DMH, FRR, MK (2) drafting or critically
revising of manuscript: AWV, PB, IKH, JWS, DMH, FRR, MK; (3) final
approval of manuscript: AWV, PB, IKH, JWS, DMH, FRR, MK.

Competing interest statement
There were no competing interests.

Funding
This work was supported by the Dutch Arthritis Foundation (grant
number 10-1-309).

Supplementary data

Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.joca.2014.05.026.

References

1. Lawrence RC, Felson DT, Helmick CG, Arnold LM, Choi H,
Deyo RA, et al. Estimates of the prevalence of arthritis and
other rheumatic conditions in the United States. Part II.
Arthritis Rheum 2008;58:26e35.

2. van Saase JL, van Romunde LK, Cats A, Vandenbroucke JP,
Valkenburg HA. Epidemiology of osteoarthritis: Zoetermeer
survey. Comparison of radiological osteoarthritis in a Dutch
population with that in 10 other populations. Ann Rheum Dis
1989;48:271e80.

3. Haugen IK, Englund M, Aliabadi P, Niu J, Clancy M, Kvien TK,
et al. Prevalence, incidence and progression of hand osteoar-
thritis in the general population: the Framingham Osteoar-
thritis Study. Ann Rheum Dis 2011;70:1581e6.

4. Mahendira D, Towheed TE. Systematic review of non-surgical
therapies for osteoarthritis of the hand: an update. Osteoar-
thritis Cartilage 2009;17:1263e8.

5. Zhang W, Doherty M, Leeb BF, Alekseeva L, Arden NK,
Bijlsma JW, et al. EULAR evidence based recommendations for
the management of hand osteoarthritis: report of a Task Force
of the EULAR Standing Committee for International Clinical
Studies Including Therapeutics (ESCISIT). Ann Rheum Dis
2007;66:377e88.

6. Towheed TE. Systematic review of therapies for osteoarthritis
of the hand. Osteoarthritis Cartilage 2005;13:455e62.
7. Bellamy N, Kirwan J, Boers M, Brooks P, Strand V, Tugwell P,
et al. Recommendations for a core set of outcome measures for
future phase III clinical trials in knee, hip, and hand osteoar-
thritis. Consensus development at OMERACT III. J Rheumatol
1997;24:799e802.

8. Altman R, Brandt K, Hochberg M, Moskowitz R, Bellamy N,
Bloch DA, et al. Design and conduct of clinical trials in patients
with osteoarthritis: recommendations from a task force of the
Osteoarthritis Research Society. Results from a workshop.
Osteoarthritis Cartilage 1996;4:217e43.

9. Maheu E, Altman RD, Bloch DA, Doherty M, Hochberg M,
Mannoni A, et al. Design and conduct of clinical trials in pa-
tients with osteoarthritis of the hand: recommendations from
a task force of the Osteoarthritis Research Society Interna-
tional. Osteoarthritis Cartilage 2006;14:303e22.

10. Kellgren JH, Lawrence JS. Radiological assessment of osteo-
arthrosis. Ann Rheum Dis 1957;16:494e502.

11. Kessler S, Dieppe P, Fuchs J, Sturmer T, Gunther KP. Assessing
the prevalence of hand osteoarthritis in epidemiological
studies. The reliability of a radiological hand scale. Ann Rheum
Dis 2000;59:289e92.

12. Kallman DA, Wigley FM, Scott Jr WW, Hochberg MC, Tobin JD.
New radiographic grading scales for osteoarthritis of the hand.
Reliability for determining prevalence and progression.
Arthritis Rheum 1989;32:1584e91.

13. Altman RD, Hochberg M, Murphy Jr WA, Wolfe F, Lequesne M.
Atlas of individual radiographic features in osteoarthritis.
Osteoarthritis Cartilage 1995;3(Suppl A):3e70.

14. Verbruggen G, Veys EM. Numerical scoring systems for the
anatomic evolution of osteoarthritis of the finger joints.
Arthritis Rheum 1996;39:308e20.

15. Verbruggen G, Wittoek R, Vander CB, Elewaut D. Morbid
anatomy of 'erosive osteoarthritis' of the interphalangeal
finger joints: an optimised scoring system to monitor
disease progression in affected joints. Ann Rheum Dis
2010;69:862e7.

16. Bijsterbosch J, Haugen IK, Malines C, Maheu E, Rosendaal FR,
Watt I, et al. Reliability, sensitivity to change and feasibility of
three radiographic scoring methods for hand osteoarthritis.
Ann Rheum Dis 2011;70:1465e7.

17. Maheu E, Cadet C, Gueneugues S, Ravaud P, Dougados M.
Reproducibility and sensitivity to change of four scoring
methods for the radiological assessment of osteoarthritis of
the hand. Ann Rheum Dis 2007;66:464e9.

18. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting
items for systematic reviews and meta-analyses: the PRISMA
statement. BMJ 2009;339:b2535.

19. Boers M, Brooks P, Strand CV, Tugwell P. The OMERACT filter
for outcome measures in Rheumatology. J Rheumatol 1998;25:
198e9.

20. Addimanda O, Mancarella L, Dolzani P, Punzi L, Fioravanti A,
Pignotti E, et al. Clinical and radiographic distribution of
structural damage in erosive and nonerosive hand osteoar-
thritis. Arthritis Care Res (Hoboken ) 2012;64:1046e53.

21. Haugen IK, Boyesen P, Slatkowsky-Christensen B, Sesseng S,
Bijsterbosch J, van der Heijde D, et al. Comparison of features
by MRI and radiographs of the interphalangeal finger joints in
patients with hand osteoarthritis. Ann Rheum Dis 2012;71:
345e50.

22. Kwok WY, Bijsterbosch J, Malm SH, Biermasz NR, Huetink K,
Nelissen RG, et al. Validity of joint space width measurements
in hand osteoarthritis. Osteoarthritis Cartilage 2011;19:
1349e55.

23. Mancarella L, Magnani M, Addimanda O, Pignotti E, Galletti S,
Meliconi R. Ultrasound-detected synovitis with power Doppler

http://dx.doi.org/10.1016/j.joca.2014.05.026
http://dx.doi.org/10.1016/j.joca.2014.05.026
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref1
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref1
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref1
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref1
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref1
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref2
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref2
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref2
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref2
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref2
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref2
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref3
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref3
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref3
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref3
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref3
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref4
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref4
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref4
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref4
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref5
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref5
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref5
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref5
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref5
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref5
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref5
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref6
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref6
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref6
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref7
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref7
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref7
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref7
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref7
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref7
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref8
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref8
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref8
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref8
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref8
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref8
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref9
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref9
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref9
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref9
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref9
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref9
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref10
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref10
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref10
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref11
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref11
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref11
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref11
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref11
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref12
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref12
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref12
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref12
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref12
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref13
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref13
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref13
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref13
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref14
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref14
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref14
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref14
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref15
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref15
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref15
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref15
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref15
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref15
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref16
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref16
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref16
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref16
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref16
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref17
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref17
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref17
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref17
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref17
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref18
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref18
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref18
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref19
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref19
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref19
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref19
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref20
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref20
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref20
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref20
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref20
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref21
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref21
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref21
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref21
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref21
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref21
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref22
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref22
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref22
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref22
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref22
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref23
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref23


A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e17231722
signal is associated with severe radiographic damage and
reduced cartilage thickness in hand osteoarthritis. Osteoar-
thritis Cartilage 2010;18:1263e8.

24. Haugen IK, Slatkowsky-Christensen B, Boyesen P, van der
Heijde D, Kvien TK. Cross-sectional and longitudinal associa-
tions between radiographic features and measures of pain and
physical function in hand osteoarthritis. Osteoarthritis Carti-
lage 2013;21:1191e8.

25. Verbruggen G, Wittoek R, Vander CB, Elewaut D. Tumour ne-
crosis factor blockade for the treatment of erosive osteoar-
thritis of the interphalangeal finger joints: a double blind,
randomised trial on structure modification. Ann Rheum Dis
2012;71:891e8.

26. Verbruggen G, Goemaere S, Veys EM. Systems to assess the
progression of finger joint osteoarthritis and the effects of
disease modifying osteoarthritis drugs. Clin Rheumatol
2002;21:231e43.

27. Botha-Scheepers S, Watt I, Breedveld FC, Kloppenburg M.
Reading radiographs in pairs or in chronological order in-
fluences radiological progression in osteoarthritis. Rheuma-
tology (Oxford) 2005;44:1452e5.

28. Botha-Scheepers S, Riyazi N, Watt I, Rosendaal FR, Slagboom E,
Bellamy N, et al. Progression of hand osteoarthritis over 2
years: a clinical and radiological follow-up study. Ann Rheum
Dis 2009;68:1260e4.

29. Botha-Scheepers SA, Watt I, Slagboom E, Meulenbelt I,
Rosendaal FR, Breedveld FC, et al. Influence of familial factors
on radiologic disease progression over two years in siblings
with osteoarthritis at multiple sites: a prospective longitudinal
cohort study. Arthritis Rheum 2007;57:626e32.

30. Buckland-Wright JC, Macfarlane DG, Lynch JA, Clark B. Quan-
titative microfocal radiographic assessment of progression in
osteoarthritis of the hand. Arthritis Rheum 1990;33:57e65.

31. Olejarova M, Kupka K, Pavelka K, Gatterova J, Stolfa J. Com-
parison of clinical, laboratory, radiographic, and scintigraphic
findings in erosive and nonerosive hand osteoarthritis. Results
of a two-year study. Jt Bone Spine 2000;67:107e12.

32. Drape JL, Idy-Peretti I, Goettmann S, Salon A, Abimelec P,
Guerin-Surville H, et al. MR imaging of digital mucoid cysts.
Radiology 1996;200:531e6.

33. Bagge E, Bjelle A, Eden S, Svanborg A. Osteoarthritis in the
elderly: clinical and radiological findings in 79 and 85 year
olds. Ann Rheum Dis 1991;50:535e9.

34. Caspi D, Flusser G, Farber I, Ribak J, Leibovitz A, Habot B, et al.
Clinical, radiologic, demographic, and occupational aspects of
hand osteoarthritis in the elderly. Semin Arthritis Rheum
2001;30:321e31.

35. Cicuttini FM, Baker J, Hart DJ, Spector TD. Relation between
Heberden's nodes and distal interphalangeal joint osteophytes
and their role as markers of generalised disease. Ann Rheum
Dis 1998;57:246e8.

36. Hart DJ, Spector TD, Brown P, Wilson P, Doyle DV, Silman AJ.
Clinical signs of early osteoarthritis: reproducibility and rela-
tion to X ray changes in 541 women in the general population.
Ann Rheum Dis 1991;50:467e70.

37. Hart D, Spector T, Egger P, Coggon D, Cooper C. Defining
osteoarthritis of the hand for epidemiological studies: the
Chingford Study. Ann Rheum Dis 1994;53:220e3.

38. Jonsson H, Helgadottir GP, Aspelund T, Sverrisdottir JE,
Eiriksdottir G, Sigurdsson S, et al. The use of digital photo-
graphs for the diagnosis of hand osteoarthritis: the AGES-
Reykjavik study. BMC Musculoskelet Disord 2012;13:20.

39. Marshall M, van der Windt D, Nicholls E, Myers H, Dziedzic K.
Radiographic thumb osteoarthritis: frequency, patterns and
associations with pain and clinical assessment findings in a
community-dwelling population. Rheumatology (Oxford)
2011;50:735e9.

40. Mathiessen A, Haugen IK, Slatkowsky-Christensen B,
Boyesen P, Kvien TK, Hammer HB. Ultrasonographic assess-
ment of osteophytes in 127 patients with hand osteoarthritis:
exploring reliability and associations with MRI, radiographs
and clinical joint findings. Ann Rheum Dis 2013;72:51e6.

41. Rees F, Doherty S, Hui M, Maciewicz R, Muir K, Zhang W, et al.
Distribution of finger nodes and their association with un-
derlying radiographic features of osteoarthritis. Arthritis Care
Res (Hoboken ) 2012;64:533e8.

42. Stern AG, Moxley G, Sudha Rao TP, Disler D, McDowell C,
Park M, et al. Utility of digital photographs of the hand for
assessing the presence of hand osteoarthritis. Osteoarthritis
Cartilage 2004;12:360e5.

43. Dahaghin S, Bierma-Zeinstra SM, Ginai AZ, Pols HA, Hazes JM,
Koes BW. Prevalence and pattern of radiographic hand oste-
oarthritis and association with pain and disability (the Rot-
terdam study). Ann Rheum Dis 2005;64:682e7.

44. Ding H, Solovieva S, Vehmas T, Riihimaki H, Leino-Arjas P.
Finger joint pain in relation to radiographic osteoarthritis and
joint locationea study of middle-aged female dentists and
teachers. Rheumatology (Oxford) 2007;46:1502e5.

45. Dominick KL, Jordan JM, Renner JB, Kraus VB. Relationship of
radiographic and clinical variables to pinch and grip strength
among individuals with osteoarthritis. Arthritis Rheum
2005;52:1424e30.

46. El-Sherif HE, Kamal R, Moawyah O. Hand osteoarthritis and
bone mineral density in postmenopausal women; clinical
relevance to hand function, pain and disability. Osteoarthritis
Cartilage 2008;16:12e7.

47. Jones G, Cooley HM, Bellamy N. A cross-sectional study of the
association between Heberden's nodes, radiographic osteoar-
thritis of the hands, grip strength, disability and pain. Osteo-
arthritis Cartilage 2001;9:606e11.

48. Kortekaas MC, Kwok WY, Reijnierse M, Huizinga TW,
Kloppenburg M. Osteophytes and joint space narrowing are
independently associated with pain in finger joints in hand
osteoarthritis. Ann Rheum Dis 2011;70:1835e7.

49. Lee HJ, Paik NJ, Lim JY, Kim KW, Gong HS. The impact of digit-
related radiographic osteoarthritis of the hand on grip-
strength and upper extremity disability. Clin Orthop Relat
Res 2012;470:2202e8.

50. Ozkan B, Keskin D, Bodur H, Barca N. The effect of radiological
hand osteoarthritis on hand function. Clin Rheumatol
2007;26:1621e5.

51. Sonne-Holm S, Jacobsen S. Osteoarthritis of the first carpo-
metacarpal joint: a study of radiology and clinical epidemi-
ology. Results from the Copenhagen Osteoarthritis Study.
Osteoarthritis Cartilage 2006;14:496e500.

52. Zhang Y, Niu J, Kelly-Hayes M, Chaisson CE, Aliabadi P,
Felson DT. Prevalence of symptomatic hand osteoarthritis and
its impact on functional status among the elderly: the Fra-
mingham Study. Am J Epidemiol 2002;156:1021e7.

53. Sunk IG, Amoyo-Minar L, Niederreiter B, Soleiman A,
Kainberger F, Smolen JS, et al. Histopathological correlation
supports the use of X-rays in the diagnosis of hand osteoar-
thritis. Ann Rheum Dis 2013;72:572e7.

54. Grainger AJ, Farrant JM, O'Connor PJ, Tan AL, Tanner S,
Emery P, et al. MR imaging of erosions in interphalangeal joint
osteoarthritis: is all osteoarthritis erosive? Skeletal Radiol
2007;36:737e45.

55. Wittoek R, Jans L, Lambrecht V, Carron P, Verstraete K,
Verbruggen G. Reliability and construct validity of ultraso-
nography of soft tissue and destructive changes in erosive

http://refhub.elsevier.com/S1063-4584(14)01108-X/sref23
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref23
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref23
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref23
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref24
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref24
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref24
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref24
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref24
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref24
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref25
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref25
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref25
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref25
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref25
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref25
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref26
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref26
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref26
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref26
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref26
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref27
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref27
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref27
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref27
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref27
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref28
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref28
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref28
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref28
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref28
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref29
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref29
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref29
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref29
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref29
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref29
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref30
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref30
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref30
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref30
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref31
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref31
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref31
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref31
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref31
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref32
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref32
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref32
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref32
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref33
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref33
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref33
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref33
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref34
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref34
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref34
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref34
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref34
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref35
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref35
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref35
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref35
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref35
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref36
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref36
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref36
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref36
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref36
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref37
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref37
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref37
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref37
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref38
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref38
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref38
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref38
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref39
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref39
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref39
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref39
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref39
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref39
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref40
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref40
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref40
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref40
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref40
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref40
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref41
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref41
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref41
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref41
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref41
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref42
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref42
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref42
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref42
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref42
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref43
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref43
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref43
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref43
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref43
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref44
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref44
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref44
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref44
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref44
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref44
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref45
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref45
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref45
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref45
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref45
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref46
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref46
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref46
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref46
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref46
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref47
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref47
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref47
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref47
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref47
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref48
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref48
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref48
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref48
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref48
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref49
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref49
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref49
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref49
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref49
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref50
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref50
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref50
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref50
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref51
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref51
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref51
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref51
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref51
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref52
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref52
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref52
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref52
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref52
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref53
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref53
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref53
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref53
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref53
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref54
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref54
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref54
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref54
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref54
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref55
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref55
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref55


A.W. Visser et al. / Osteoarthritis and Cartilage 22 (2014) 1710e1723 1723
osteoarthritis of the interphalangeal finger joints: a compari-
son with MRI. Ann Rheum Dis 2011;70:278e83.

56. Iagnocco A, Filippucci E, Ossandon A, Ciapetti A, Salaffi F,
Basili S, et al. High resolution ultrasonography in detection of
bone erosions in patients with hand osteoarthritis.
J Rheumatol 2005;32:2381e3.

57. Keen HI, Wakefield RJ, Grainger AJ, Hensor EM, Emery P,
Conaghan PG. Can ultrasonography improve on radiographic
assessment in osteoarthritis of the hands? A comparison be-
tween radiographic and ultrasonographic detected pathology.
Ann Rheum Dis 2008;67:1116e20.

58. Vlychou M, Koutroumpas A, Malizos K, Sakkas LI. Ultrasono-
graphic evidence of inflammation is frequent in hands of pa-
tients with erosive osteoarthritis. Osteoarthritis Cartilage
2009;17:1283e7.

59. Huetink K, van't Klooster R, Kaptein BL, Watt I, Kloppenburg M,
Nelissen RG, et al. Automatic radiographic quantification of
hand osteoarthritis; accuracy and sensitivity to change in joint
space width in a phantom and cadaver study. Skelet Radiol
2012;41:41e9.

60. van't Klooster R, Hendriks EA, Watt I, Kloppenburg M,
Reiber JH, Stoel BC. Automatic quantification of osteoarthritis
in hand radiographs: validation of a new method to measure
joint space width. Osteoarthritis Cartilage 2008;16:18e25.

61. Saltzherr MS, van Neck JW, Muradin GS, Ouwendijk R,
Luime JJ, Coert JH, et al. Computed tomography for the
detection of thumb base osteoarthritis: comparison with dig-
ital radiography. Skelet Radiol 2013;42:715e21.

62. Ceceli E, Gul S, Borman P, Uysal SR, Okumus M. Hand function
in female patients with hand osteoarthritis: relation with
radiological progression. Hand (NY) 2012;7:335e40.

63. Marshall M, van der Windt D, Nicholls E, Myers H, Hay E,
Dziedzic K. Radiographic hand osteoarthritis: patterns and
associations with hand pain and function in a community-
dwelling sample. Osteoarthritis Cartilage 2009;17:1440e7.

64. Buckland-Wright JC, Macfarlane DG, Lynch JA. Sensitivity of
radiographic features and specificity of scintigraphic imaging
in hand osteoarthritis. Rev Rhum Engl Ed 1995;62:14Se26S.

65. Marshall M, Dziedzic KS, van der Windt DA, Hay EM.
A systematic search and narrative review of radiographic
definitions of hand osteoarthritis in population-based studies.
Osteoarthritis Cartilage 2008;16:219e26.

66. Kerkhof HJ, Meulenbelt I, Akune T, Arden NK, Aromaa A,
Bierma-Zeinstra SM, et al. Recommendations for standardiza-
tion and phenotype definitions in genetic studies of osteoar-
thritis: the TREAT-OA consortium. Osteoarthritis Cartilage
2011;19:254e64.

67. Haugen IK, Boyesen P. Imaging modalities in hand osteo-
arthritiseand perspectives of conventional radiography,
magnetic resonance imaging, and ultrasonography. Arthritis
Res Ther 2011;13:248.

68. Spaans AJ, van Laarhoven CM, Schuurman AH, van Minnen LP.
Interobserver agreement of the Eaton-Littler classification
system and treatment strategy of thumb carpometacarpal
joint osteoarthritis. J Hand Surg Am 2011;36:1467e70.

69. Sunk IG, Amoyo-Minar L, Stamm T, Haider S, Niederreiter B,
Supp G, et al. Interphalangeal Osteoarthritis Radiographic
Simplified (iOARS) score: a radiographic method to detect
osteoarthritis of the interphalangeal finger joints based on its
histopathological alterations. Ann Rheum Dis 2013 [epub
ahead of print].

70. Altman RD, Fries JF, Bloch DA, Carstens J, Cooke TD, Genant H,
et al. Radiographic assessment of progression in osteoarthritis.
Arthritis Rheum 1987;30:1214e25.

71. Paradowski PT, Lohmander LS, Englund M. Natural history of
radiographic features of hand osteoarthritis over 10 years.
Osteoarthritis Cartilage 2010;18:917e22.

72. Maheu E, Cadet C, Carrat F, Barthe Y, Berenbaum F. Radiologic
Progression of Hand Osteoarthritis Over 2.6 Years According to
Various Methods of Calculation e Data From the SEKOIA Trial
2013.

73. Dela Rosa TL, Vance MC, Stern PJ. Radiographic optimization of
the Eaton classification. J Hand Surg Br 2004;29:173e7.

74. Burnett S, Hart DJ, Cooper C, Spector TD. A Radiographic Atlas
of Osteoarthritis 1994.

75. Eaton RG, Glickel SZ. Trapeziometacarpal osteoarthritis. Stag-
ing as a rationale for treatment. Hand Clin 1987;3:455e71.

76. Lane NE, Nevitt MC, Genant HK, Hochberg MC. Reliability of
new indices of radiographic osteoarthritis of the hand and hip
and lumbar disc degeneration. J Rheumatol 1993;20:1911e8.

http://refhub.elsevier.com/S1063-4584(14)01108-X/sref55
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref55
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref55
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref56
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref56
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref56
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref56
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref56
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref57
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref57
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref57
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref57
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref57
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref57
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref58
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref58
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref58
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref58
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref58
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref59
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref59
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref59
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref59
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref59
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref59
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref60
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref60
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref60
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref60
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref60
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref61
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref61
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref61
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref61
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref61
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref62
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref62
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref62
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref62
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref63
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref63
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref63
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref63
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref63
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref64
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref64
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref64
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref64
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref66
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref66
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref66
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref66
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref66
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref67
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref67
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref67
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref67
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref67
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref67
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref68
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref68
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref68
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref68
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref68
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref69
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref69
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref69
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref69
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref69
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref70
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref70
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref70
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref70
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref70
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref70
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref71
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref71
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref71
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref71
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref72
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref72
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref72
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref72
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref73
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref73
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref73
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref73
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref73
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref74
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref74
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref74
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref75
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref75
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref76
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref76
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref76
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref77
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref77
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref77
http://refhub.elsevier.com/S1063-4584(14)01108-X/sref77

	Radiographic scoring methods in hand osteoarthritis – a systematic literature search and descriptive review
	Introduction
	Methods
	Identification of studies
	Inclusion and exclusion criteria
	Data extraction
	Statistical analyses

	Results
	Literature flow
	Study characteristics
	Discrimination
	Reliability
	Sensitivity to change

	Feasibility
	Validity

	Discussion
	Contributions
	Competing interest statement
	Funding
	Appendix A Supplementary data
	References