Rheumatology 2005;44:577586 doi:10.

1093/rheumatology/keh549
Advance Access publication 3 February 2005
Review

Patient-assessed health in ankylosing spondylitis:

a structured review
K. L. Haywood, A. M. Garratt1 and P. T. Dawes2

Objective. To review evidence relating to the measurement properties for all disease-specic, multi-item, patient-assessed health
instruments in patients with ankylosing spondylitis (AS).
Methods. Systematic literature searches were made to identify instruments, using predened criteria relating to reliability,
validity, responsiveness and precision.
Results. Twelve AS-specic and three arthritis-specic instruments met the inclusion criteria. Three AS-specic instruments
that measure health-related quality of life (HRQL) were reviewed. The Bath Ankylosing Spondylitis Disease Activity Index
(BASDAI), the Bath Ankylosing Spondylitis Functional Index (BASFI) and the Dougados Functional Index (DFI) had the
greatest amount of evidence for reliability, validity and responsiveness across a range of settings. Four instruments lacked
evidence for testretest or internal consistency reliability. Most were assessed for validity through comparisons with other
instruments, global judgements of health, mobility or clinical and sociodemographic variables. Most were assessed for
responsiveness through mean score changes. Three instruments lacked evidence of responsiveness.
Conclusion. This review provides a contribution to AS assessment. AS-specic multi-item measures specic to the assessment of
pain, stiffness, fatigue and global health were not identied; where assessed, these domains were largely measured with single-
item visual analogue scales. Single items may provide a limited reection of these important domains. The BASFI and DFI
remain the instruments of choice for functional assessment. HRQL is recommended as a core assessment domain. Further
concurrent evaluation is recommended.
KEY WORDS: Ankylosing spondylitis, Patient-assessed health, Measuring properties.

Ankylosing spondylitis (AS) is an incurable, inammatory disease,
primarily affecting the pelvis and spine, but often with involvement
of peripheral joints, entheses and extra-articular sites [1]. Often
affecting individuals from an early adult age, the disease can have
a profound impact on life quality in terms of physical, social and
psychological well-being [2]. It is widely accepted that the patients
perception of disease impact and the outcomes of health-care
should be included in clinical trials and similar forms of evaluative
study. This has resulted in a signicant increase in the availability
of patient-assessed health instruments which aim to measure
aspects of health from the perspective of the patient [3].
Structured reviews of measurement properties and accompanying
professional consensus are helpful in supporting instrument
selection and standardization [3, 4].
To encompass the multidimensional impact of AS, ve core
evaluative domains and, with the exception of functional disability,
associated single-item scales have been recommended by the
Assessment in AS International Working Group (ASAS; www.
asas-group.org): pain [intensity; visual analogue scale (VAS)],
spinal stiffness/inammation (morning stiffness duration; VAS),
functional ability [Bath Ankylosing Spondylitis Functional Index
(BASFI) [5] or Dougados Functional Index (DFI) [6]], patient
global assessment of health status (VAS; last week) and spinal
mobility [7, 8]. Instrument selection was informed by a literature
review (19881995; Medline database and citation searches),
expert opinion and professional consensus agreement.
Single items, such as those recommended by ASAS, may not
allow patients to appropriately report the wide impact of disease or
treatment, providing a limited reection of health [4]. The resulting
summary judgement of health status limits measurement validity
and score interpretation [4, 9]. Patient-assessed health instruments
usually take the form of questionnaires containing multiple items,
or questions, to reect the broad nature of health status, disease
or injury [3, 4]. These instruments aim to provide an accurate
assessment of health or disease from the patients perspective,
which contribute to validity and score interpretation [3, 4, 9]. This
structured review presents an updated and more extensive review
of published evidence for AS- and arthritis-specic multi-item,
patient-assessed instruments that measure any aspect of health or
health-related quality of life (HRQL) in this disease (19882004).
The review will inform instrument selection and future research
within this eld.
Methods
Identication of studies
The search strategy was designed to retrieve references relating
to the development and evaluation of multi-item patient-assessed

Unit of Health-Care Epidemiology, Department of Public Health, University of Oxford, Oxford, 1Staffordshire Rheumatology Centre, University Hospital
of North Staffordshire NHS Trust, Stoke-on-Trent, UK and 2Norwegian Centre for Health Services Research, St Olavs Plass, Oslo, Norway.

Submitted 14 September 2004; revised version accepted 24 December 2004.

Correspondence to: K. Haywood, Unit of Health-Care Epidemiology, Department of Public Health, University of Oxford, Oxford OX3 7LF, UK.
E-mail: kirstie.haywood@uhce.ox.ac.uk
577
The Author 2005. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oupjournals.org
578 K. L. Haywood et al.
health instruments, including reviews. The rst AS-specic patient-
assessed health instrument was developed in 1988 [6], so the search
strategy was restricted to the period from 1988 to August 2004.
All searches included terms specic to AS combined with the
measurement of health outcome and instrument measurement
properties [4, 10]. Search strategies used medical subject headings
(MeSH terms) and free text searching. Further searches used
names of identied instruments.
Databases searched included AMED, CINAHL, the Cochrane
Controlled Trials Register, EMBASE, Medline and Psychlit. The
Patient-assessed Health Instruments (PHI) bibliographic data-
base (http://phi.uhce.ox.ac.uk/), hosted by the National Centre for
Health Outcomes Development at the University of Oxford, was
also searched. This database is based on systematic searches of
the literature and contains over 7000 records relating to published
instrument evaluations found on the major electronic databases
(from database inception to September 2003) [3].
The reference lists of included articles were reviewed for
additional articles. Relevant journals were hand-searched, includ-
ing Annals of the Rheumatic Diseases, Arthritis and Rheumatism,
Rheumatology and the Journal of Rheumatology. Texts and
compendia were consulted [1113]. The reference lists of exist-
ing reviews [8, 1417] and manuscripts discussing the clinical
management of AS [18, 19] were also reviewed.
Inclusion criteria
Titles and abstracts of all articles were assessed for inclusion/
exclusion by two independent reviewers (K.L.H., A.M.G.) and
agreement was checked. Published articles were included if they
provided evidence of measurement properties for AS or arthritis-
specic, multi-item, patient-assessed health instruments following
completion by adults with AS. Generic or domain-specic multi-
item instruments (not specic to AS or arthritis) were excluded.
Clinician-assessed instruments, single-item and mobility measures,
radiographic and imaging techniques were excluded. The review
included instrument evaluations in non-English-speaking popula-
tions that were published in an English language journal.
Instruments without evidence of reliability or validity were
excluded.
Data extraction
Data extraction followed predened criteria considered important
in the evaluation of patient-assessed health instruments [4, 9, 11]
and included patient characteristics, type of instrument, the
domain focus, scaling, length, and evidence of measurement
properties. The summary of evidence follows that of previous
reviews [10, 11].
Evidence for measurement properties was assessed using
accepted criteria [4, 9, 11]. Reliability assesses measurement
stability over time, and for multi-item instruments, internal
consistency [4, 9]. Testretest reliability assesses score temporal
stability and is assessed following instrument completion at two
time points; it assumes no change in underlying condition [4].
Internal consistency reliability assesses the ability of items to
measure a single underlying domain, and is assessed following a
single application. Evidence for testretest and internal consistency
reliability, specically testretest correlation coefcients and
Cronbachs , are presented. The reliability estimate reects two
components: a true score and an underlying level of error [9].
Reliability estimates range between 0 and 1.0; the closer the score
to 1.0, the lower the error [4, 9]. The reliability of an instrument has
implications for whether it is suitable for application in group or
individual evaluation. For the evaluation of individuals high levels
of reliability, above 0.90, have been recommended [4, 9]. For group
comparisons, levels over 0.70 are recommended [9].
Validity assesses whether an instrument measures what is
intended [4], and is referred to as truth within the OMERACT
TABLE 1. Grading scale for summary of evidence for instrument reliability,
validity and responsiveness (adapted from McDowell and Newell [11])
Thoroughness of evaluation Results of evaluation
0 No published evidence 0 No published
numerical results
Basic information only Weak evidence only
Several types of test, or Moderate evidence
several evaluations in
different populations
All major forms of validity/ Strong evidence
reliability/responsiveness
reported. Several good
quality evaluations in
different populations
lter [20]. Content and face validity are assessed through an
appraisal of item content; evidence for the source of instrument
items. Evidence for these forms of qualitative validity is presented.
Evidence for external construct validity requires comparison of
instrument scores with those for other measures of health, clinical,
socio-demographic and health service use variables [4, 10].
Construct validity may also be assessed by group or divergent
validity, where, based on theory or existing evidence, we can
state that one group will possess more or less of a construct [9].
For example, compared with the general population, people with
AS may be expected to report greater levels of pain and worse
HRQL. The results of these comparisons are presented. Evidence
derived from statistical methods such as factor analysis to
describe instrument dimensionality or internal construct validity
is presented [4].
Responsiveness has been described as the ability of an instru-
ment to measure clinically important change over time when
change is present [4], and is a necessary property of instruments
intended for application in evaluative studies [21, 22]. Two broad
approaches to evaluating responsiveness include those that are
distribution-based and those that are anchor-based. Distribution-
based approaches relate changes in instrument score to some
measure of variability, the most common being the effect size
statistic [4]. Effect size statistics provide a standardized unit of
expression of the size and meaning of score change, supporting
the comparison of instrument performance; the most responsive
instruments have larger effect sizes. Anchor-based approaches
assess the relationship between changes in instrument score and
an external variable [23]. This includes health transition items or
global judgements of change. Responses to transition items have
also been compared to instrument score change using correlation.
Data extraction covered the full range of approaches to measuring
responsiveness and included descriptive statistics.
Data summary
The summary of reviewed evidence was informed by previous
instrument reviews [11, 24, 25] (Table 1). The thoroughness (that is,
the range of evaluations) and results of evaluations are considered
separately. A summary scale from 0 to is presented, where 0
is no evidence for the underlying criteria and indicates a wide
range of testing and good evidence of measurement properties
(Table 1).
Results
Identication of studies
The literature searches produced 138 articles covering 12
AS-specic and three arthritis-specic multi-item, patient-assessed
health instruments with evidence of reliability or validity following
completion by patients with AS (Table 2).
TABLE 2. AS-specic patient-assessed health instruments

Instrument Developer Domains (items) Response scale Score Origin Translations

Symptoms and disease activity
Bath AS Disease Activity Garrett et al. (1994) [28] Disease activity (6) VAS; adjectival anchors 010; 0 disease activity UK Multiple
Index (BASDAI)
Body Chart Dziedzic (1997) [29] Pain (1) 4-point descriptive No max score; 0 no pain UK

Function

Patient-assessed health in AS: a review

ASAQ Nemeth et al. (1987) [37] Function/spinal mobility (2) 8-point adjectival 011; 0 best function UK
BASFI Calin et al. (1994) [5] Function (10) VAS; adjectival anchors 010; 0 best function UK Multiple
DFI Dougados et al. (1988) [6] Function (20) 3-point categorical 040; 0 best function France Multiple
(revised to 5-point)
HAQa Fries (1980) 32 Arthritis-specic (core): 03; 0 best function USA Multiple
Disability (20), 4-point categorical; 03; 0 no pain
Discomfort (1) 1
15 cm VAS
HAQ-S Daltroy et al. (1990) [33] (HAQ AS-specic items) 03; 0 best function USA/UK Dutch
Disability (23), 4-point categorical; 03; 0 no pain
Discomfort (2) 2
15 cm VAS
RLDQ Abbott et al. (1994) [38] Function (16) 4-point categorical 048; 0 best function UK Swedish

Global well-being
BAS-G Jones et al. (1996) [30] Global well-being (2) VAS; adjectival anchors 010; 0 best well-being UK

Health-related quality of life

ASQoL Doward et al. (2003) [39] HRQL (18) Yes/No 018; 0 best HRQL UK, NL Dutch
AIMSa Meenan et al. (1980) [40] 9 domains (45) 26 point categorical 010; 0 best health USA Multiple
AIMS2a Meenan et al. (1992) [41] 12 domains (57) 5 point categorical 010; 0 best health USA
AS-AIMS2 Guillemin et al. (1999) [31] 13 domains (63) 5 point categorical 010; 0 best health France
PETb Bakker et al. (1995) [34] Single index (15) 7-point descriptive 049; 0 best health Canada/NL
PGI Haywood et al. (2003) [36] Single index (7) 10-point descriptive 010; 10 best HRQL UK
plus weighting
a
Arthritis-specic instruments; binterview-administered.

579
580 K. L. Haywood et al.

Patient characteristics AS- or arthritis-specic, multi-item patient-assessed measures

The number of respondents ranged from 14 [26] to 4282 [27]. All
patients had a clinical diagnosis of AS. Mean respondent ages
ranged from 30 to 57 yr. Mean disease duration ranged from 5.5 to
32.4 yr.
specic to the assessment of pain, stiffness or fatigue were not
identied; where assessed, these domains were largely measured
with single-item visual analogue scales.

Patient-assessed health instruments
The Bath Ankylosing Spondylitis Disease Activity Index
(BASDAI) [28], BASFI [5] and DFI [6] have undergone the
greatest amount of testing for all measurement properties, with
72, 70 and 46 published articles respectively (Table 3).
The shortest instruments were the Body Chart [29] and the Bath
Ankylosing Spondylitis Global score (BAS-G) [30], with one and
two items respectively, although a single item is often reported for
the BAS-G (Table 1). The longest was the AS-specic Arthritis
Impact Measurement Scale (63 items) [31]. Most instruments
produce index scores; that is, item scores are summed to produce
a single score. The Health Assessment Questionnaire (HAQ) [32]
and the HAQ-S [33] produce prole scores; that is, item scores are
summed within separate domains, providing a reection of health
across these domains. The Patient Elicitation Technique (PET)
[34, 35] and Patient-Generated Index (PGI) [36] are individualized
measures. With the exception of the PET, which requires interview
administration [35], all instruments have been self-completed.
Instruments are grouped in Tables 2 and 3 according to the
domains that they purport to measure: symptoms and disease
activity (BASDAI, Body Chart); function [Assessment in
Ankylosing Spondylitis Questionnaire (ASAQ) [37], BASFI [5],
DFI [6], HAQ [32], HAQ-S [33] and the Revised Leeds Disability
Questionnaire (RLDQ) [38]]; global well-being (BAS-G [30]); and
HRQL (ASQoL [39], AIMS [40], AIMS2 [41], AS-AIMS2 [31],
PET [35] and PGI [36]).
Reliability
The BASDAI and BASFI have the greatest evidence of reliability
(Table 2). Seven instruments have evidence of internal consistency
reliability: BASDAI, BASFI, DFI, HAQ-S, RLDQ, ASQoL and
AS-AIMS2. Alpha levels for studies evaluating the English BASFI
[42], RLDQ [38, 43], ASQoL [39, 43] and a range of instrument
translations (for example, the Turkish BASFI [44], Dutch [45]
and Finnish versions of the DFI [46]) exceeds 0.90, the criterion
recommended for individual patients [4, 9]. Evidence for the
BASDAI [42, 43, 4749] and the Austrian HAQ-S [50] exceed 0.70,
the level recommended for groups [4]. Evidence of item-total
correlation for the BASDAI, RLDQ and the ASQoL support scale
homogeneity [9]. Tests of internal consistency are not appropriate
for the instruments that are not based on summated rating scales:
Body Chart, BAS-G, PET and PGI.
Ten instruments have evidence of testretest reliability:
BASDAI, Body Chart, ASAQ, BASDAI, DFI, RLDQ, BAS-G,
ASQoL, AS-AIMS2 and PGI. All reliability estimates for the
RLDQ [38, 43] and the ASQoL [39, 43] exceed criteria necessary
for individual assessment. Several reliability estimates for the
BASFI [46, 51, 52] and the DFI [5, 45, 46, 53] exceed the criteria
necessary for individuals. Most estimates for remaining instru-
ments exceed the criteria necessary for groups [9]. Low levels of
testretest reliability have been reported for the BASDAI (range
0.530.64) [54]. The retest period ranged from 1 day to 6 weeks.
Few authors indicate if reliability is assessed in patients reporting
no change in health over the retest period.

TABLE 3. Summary of evaluations for AS- and arthritis-specic patient-assessed health instruments

Published evaluations (n)a Summary of measurement propertiesc

Responsivenessb Reliability Validity Responsiveness

Instrument Total Reliability Validity PT (UC) DT Thoroughness Results Thoroughness Results Thoroughness Results
Symptoms and disease activity
BASDAI 72 17 37 15 (3) 25
Body Chart 2 2 2 0 (2) 0
Function
ASAQ 6 1 6 0 0 0 0
BASFI 70 19 29 18 26
DFI 46 16 18 15 (1) 14
HAQ 5 0 3 1 (1) 2 0 0
HAQ-S 22 4 13 10 (2) 2
RLDQ 7 6 6 3 (1) 0
Global well-being
BAS-G 18 4 9 8 4
Health-related quality of life
ASQoL 10 4 6 2 (3) 2
AIMS 8 0 7 2 (1) 0 0 0
AIMS2 3 0 1 0 0 0 0 0 0
AS-AIMS2 1 1 1 0 0 0 0
PET 1 0 1 1 0 0 0
PGI 1 1 1 0 (1) 0
a
Total number of published instrument evaluations following completion by patients with AS.
b
PT, physiotherapy intervention; UC, usual care; DT, drug therapy.
c
Summary of evidence (after McDowell and Newell, 1996 [11]). Thoroughness: 0, no published evidence; , basic information; , several types
of test or several evaluations in different populations; , all major forms of reliability/validity/responsiveness reported. Several good-quality
evaluations in different populations. Results: 0, no published results; , weak evidence; , moderate evidence; , strong evidence of
reliability/validity/responsiveness across a range of settings/interventions.
 Patient-assessed health in AS: a review
Four instruments have evidence of both forms of reliability
BASDAI, BASFI, DFI and RLDQwhere reliability estimates
support application at the group and, in some instances, the
individual level. One-week testretest reliability and associated
95% limits of agreement were calculated for the BASDAI, BASFI,
BAS-G and DFI [51]. High reliability estimates were associated
with wide score ranges; this was interpreted as indicating poor
instrument reliability [51].
Four instruments do not have evidence of reliability in patients
with AS: HAQ, AIMS, AIMS2 and PET.
Validity
The ASQoL [39], PET [34] and PGI [36] involved patients
in item generation. The BASDAI [28], BASFI [6], RLDQ [38]
and AS-AIMS [31] incorporated the opinion of health-care
professionals and patients but the role of patients is not made
explicit. The DFI included three rheumatologists and the HAQ-S
a survey of functional disability in patients with AS. With
the exception of the ASAQ [37] and BAS-G [30], for which
item generation is not described, the literature and existing
instruments provided the major source of items for the remaining
instruments.
Evidence supports the internal construct validity of ve
instruments, the results of which support the single domains
of the BASDAI [43, 49], DFI [6], RLDQ [43] and ASQoL
[43]. Although the developers of the AS-AIMS2 describe 13
domains, 12 domains were found following principal component
analysis [31].
All instruments have evidence for validity through comparison
with instruments that measure similar or related constructs, and/or
with measures of mobility. This is most extensive for the BASDAI,
BASFI and DFI (Table 3). With the exception of the HAQ, AIMS,
AS-AIMS2 and PET, all instruments have evidence to support
their ability to discriminate between groups of patients with AS
dened by clinical, radiographic and socio-demographic variables.
This is most extensive for the BASDAI, BASFI and DFI. The
BASDAI, BASFI, DFI and HAQ-S have evidence to support
their ability to discriminate between groups dened by health
service use.
Responsiveness
With the exception of the ASAQ, AIMS2 and AS-AIMS,
all instruments have some evidence of responsiveness following
completion by patients with AS (Table 3). With the exception of
the Body Chart, RLDQ, AIMS, PET and PGI, all instruments
have some evidence of responsiveness following both drug therapy
and physical therapy. This is most extensive for the BASDAI and
BASFI. Effect size statistics were reported for all instruments.
Correlation of change scores with change in other variables was
reported for the BASDAI, BASFI, DFI, HAQ-S, BAS-G, ASQoL
and AIMS. With the exception of the AIMS, Body Chart, RLDQ,
PET and PGI, seven instruments had evidence of group discrimi-
nation over time. Statistical signicance was frequently reported
but the clinical signicance of score change was rarely addressed.
There is limited evidence for the Body Chart, HAQ, HAQ-S,
RLDQ and instruments that measure HRQL. Five studies
reporting evidence for the HAQ-S describe different stages in a
trial of physical therapy [34, 5558].
Moderate to strong levels of responsiveness were reported for
the BASDAI [for example, 49, 5963] and BASFI (for example,
49, 53, 5966) following a range of placebo-controlled trials and
the longitudinal evaluation of active drugs. Mean score changes
greater than 2.0 (scale 010) were reported for both the BASDAI
and BASFI following the evaluation of anti-TNF therapy with
6- to 52-week follow-up periods.
581
Few studies reported effect size statistics for the evaluation
of instrument responsiveness following physical therapy; most
reported mean score change. Moderate and small effect sizes were
reported for the BASDAI and BASFI respectively following the
longitudinal evaluation of in-patient rehabilitation [67]. Small
effect sizes were reported for the BASDAI and BASFI following
combined spa and exercise therapy [68, 69]. The BASDAI was
responsive to improvement or deterioration in health following the
evaluation of usual care [43]. Mean score change for the BASDAI
[28, 54, 6775] and BASFI [5, 54, 6770, 7274, 76, 77] did not
exceed 1.9 and 1.3 respectively following all physical therapy
interventions within a 2- to 40-week follow-up period. BASFI
score change of less than 0.7 (scale 010) was reported following
similar in-patient rehabilitation programmes of 3 weeks duration
[67, 72, 74]; score change of less than 0.6 was reported following
a 6-week out-patient exercise programme [78]. Non-statistically
signicant score change was reported following a long-term,
prospective evaluation of function [72, 75].
Moderate to strong levels of responsiveness were reported for
the DFI following a range of drug therapy evaluations. Lower
levels of responsiveness and poor group discrimination have
been reported following physical therapy [5, 67]. There is limited
evidence of responsiveness for a modied ve-point response scale
[54, 68, 74].
The HAQ has limited evidence of responsiveness following drug
therapy evaluation in patients with AS but large score changes
were reported following the longitudinal evaluation of Iniximab
in AS [79]. Evidence suggests that the HAQ-S is not responsive to
change in functional ability following physical therapy [34, 55, 58,
68, 80].
Larger score changes have been reported for the single BAS-G
items following drug therapy evaluation [79, 82] than following
physical therapy [67, 73]. Similarly, the ASQoL was more
responsive to change following drug therapy (etanercept) [83, 84]
than following physical therapy [68, 69] or usual care [43].
Precision
Floor effects were reported for the HAQ-S (25% scored 0) [56]
and RLDQ [43]. Skewed score distributions were also reported
for the DFI [16, 85]. Floor effects may be a function of limited
item content and/or response options [4], and limit measure-
ment discrimination and responsiveness. Normal score distribu-
tions were reported for the ASQoL, BASDAI, BASFI and
PGI. Score distributions were not reported for the remaining
instruments.
Discussion
To provide the most effective management in the care of
individuals with AS it is important to determine how the disease
and treatment affect health from the patients perspective. The
application of patient-assessed health instruments has become
increasingly important within the assessment of health-care [3, 4]
and more specically within rheumatology [3, 86]. Signicant
progress in the eld has been made since the initial ASAS
recommendations, which acknowledged that they could change
following new research evidence [7, 8]. This review provides a
timely assessment of these recommendations and the rst detailed
review of AS- and arthritis-specic multi-item, patient-assessed
health instruments. The review will inform the appropriate
selection of multi-item, patient-assessed health instruments to be
used in clinical practice and research.
ASAS recommended ve core assessment domains: pain,
stiffness, function, global well-being and spinal mobility. From
the 15 reviewed multi-item instruments, one assessed pain intensity
(Body Chart); ve AS-specic instruments (ASAQ, BASFI, DFI,
582 K. L. Haywood et al.
HAQ-S and RLDQ) and one arthritis-specic (HAQ) instrument
assessed functional ability; and one instrument assessed global
well-being (BAS-G). The BASDAI, an AS-specic measure of
disease activity, and three AS-specic measures of HRQL
(AS-AIMS2, ASQoL, PGI) include items to assess pain, stiffness
and fatigue.
Pain, stiffness and fatigue are frequently described as impor-
tant symptoms by patients [36, 87] and clinicians [68]. The Body
Chart had poor evidence of measurement properties and is not
recommended without further evaluation. Although evidence of
completion rates were mixed, measurement properties support
consideration of the BASDAI as a measure of disease activity.
However, other important issues, such as item content and
response format, should also be considered.
Six measures of functional ability were reviewed. The ASAQ,
HAQ and HAQ-S have limited evidence of measurement
properties following completion by patients with AS; the HAQ
and HAQ-S also have evidence of poor data quality. The RLDQ
has good completion rates and a very high level of reliability, but
it is not recommended for application due to poor data quality,
the limited range of functional disability assessed, and limited
responsiveness.
Both the BASFI and DFI have acceptable levels of reliability.
Although both have good evidence for construct validity and
similar levels of responsiveness following drug therapy evalua-
tion, the BASFI has better content validity and is more
responsive following physical therapy. Further, evidence suggests
that the BASFI is more responsive than the DFI in the early
stages of treatment with anti-TNF therapies; comparable levels
of responsiveness were found after 4 months of treatment [66].
This may be a function of the limited response options for the
original DFI. The modied categorical response scale [74, 85]
may improve responsiveness. However, patients often experience
difculty in completing VAS, the format of the BASDAI, BASFI
and BAS-G, and reservations have been expressed about the
interpretation, acceptability and feasibility of VAS scales [4, 9, 54].
Both patients [4, 11, 54] and clinicians [88, 89] have expressed
preferences for categorical rating scales. A comparative evalua-
tion of response formats by patients with AS supported a
preference for categorical rating scales (49%), followed by
numerical rating scales (38%) and visual analogue scales (9%)
[54]. The initial ASAS recommendations for functional assess-
ment suggested use of the BASFI or DFI. Evidence suggests
that the BASFI and DFI differ quite broadly in terms of item
content, response format, levels of patient (and clinician)
acceptability and measurement properties in different settings.
This suggests that reports of functional ability from these two
instruments may not be directly comparable. Further consensus
is required to recommend a single instrument for AS-specic
functional assessment.
Evidence supports the ASAS recommendation of the BAS-G as
a single-item assessment of global well-being [8]. However, single
items provide a limited assessment of global health [4]. Since 1999,
three AS-specic measures of HRQL have been identied: ASQoL,
AS-AIMS2 and PGI. Three arthritis-specic measures of HRQL
were also reviewed: AIMS, AIMS2 and PET. These lack evidence
of reliability in AS, have limited evidence of validity and cannot be
recommended for application in AS assessment without further
evaluation. Patients were explicitly involved in item generation for
the ASQoL, PET and PGI. Patient participation enhances content
validity [4]. Although overlap between ASQoL items and PGI
areas was expected, a concurrent evaluation found that several of
the areas most frequently nominated by patients completing the
PGI (body image, walking and work outside the home) are not
addressed by items within the ASQoL [36]. The ASQoL purports
to measure AS-related quality of life but does not include some
of the 10 most important and frequently mentioned patient
concerns [36].
The ASQoL has good completion rates, satisfactory data
quality and scaling assumptions, high reliability, some evidence
of validity, and good evidence of responsiveness following
drug therapy, but small to moderate levels following physical
therapy [68]. The ASQoL has a dichotomous response scale
which often fails to support sufciently detailed descriptions
of health [4, 90, 91]. The PGI and AS-AIMS2 are two new
measures of HRQL. As a result, both have limited evidence
for their measurement properties and further evaluation is
recommended.
Instrument measurement properties are context-specic
attributes that can differ across populations, settings and inter-
ventions [21, 22, 91]. ASAS identied three treatment settings:
disease-controlling anti-rheumatic therapy (DC-ART), symptom-
modifying antirheumatic drugs (SMARDs) and physical therapy,
and clinical record keeping [7, 8]. Evidence for measurement
and practical properties across these settings should be considered
alongside item content and appropriateness to the clinical or
research question when selecting an instrument. Although ASAS
made several instrument recommendations, the majority of these
were for single-item scales which may not provide an adequate
assessment of health [4, 9]. Patient-assessed health instruments
should provide an accurate assessment of disease impact and
health-care from the patients perspective [3]. The inclusion of
multiple items within a questionnaire will provide a more detailed
and informative assessment of the health domain or concept [4].
Although quick to complete, single items may be a poor surrogate
for a patients perception of disease impact; content validity
is likely to be lower and important information may be lost
which hinders data interpretation and usefulness to clinical
decision-making [4, 9].
Low levels of testretest reliability, which did not support
application in group assessment, have been reported for a domain-
specic, multi-item measure of fatigue [Multidimensional Fatigue
Inventory (MFI)] and a single-item VAS for fatigue severity in
patients with AS [92]. MFI domains ranged from 0.57 (physical
fatigue) to 0.75 (reduced motivation; mental fatigue); fatigue
VAS 0.60. Although associated with moderate to good levels of
responsiveness, further evaluation of measurement properties is
required before either instrument can be recommended for use.
Further concurrent evaluations between AS-specic and domain-
specic, multi-item, measures of pain, fatigue and stiffness
are recommended. Where appropriate, renement of existing
instruments is required before the development of new instru-
ments; seeking the views of people with AS with regard to
instrument format, relevance and mode of completion is strongly
recommended [9].
Although a range of comparative evaluations exists, further
comparative evaluations are required of multi-item AS-specic
instruments, such as the modied DFI and BASFI, the newly
developed measures of HRQL, and widely-used generic instru-
ments, across different treatment settings. Particular attention
should be paid to the evaluation of instrument responsiveness over
longer periods, score interpretation and the role of patient-assessed
health instruments in clinical decision-making and communicating
treatment benet. Recent quality improvement initiatives within
the NHS recognize the importance of evaluating health-care
organizations and technologies in terms of patient-assessed
outcomes. For instance, the measurement of HRQL is central
to the NICE technology appraisal process. With the increasing
availability of new and expensive therapies in rheumatology, the
challenge to provide more informative, responsive and relevant
patient-based assessment of disease impact and treatment efcacy
becomes ever more important.
All reviewed studies reported instrument evaluations following
completion by groups of patients participating in clinical trials
or longitudinal evaluations of care. Although widely accepted in
clinical trials [4], there is currently little evidence to support the
effectiveness of including patient-assessed health instruments in
 Patient-assessed health in AS: a review
routine practice and clinical record keeping [93]. Most studies
provide group-level estimates of instrument performance and score
change, the interpretation of which may be difcult for clinicians
wishing to use patient-assessed instruments to inform clinical
decision-making at the individual level [21]. Further exploration
of the acceptability, feasibility and interpretation of including
patient-assessed health instruments in clinical practice and of the
benet to be gained is required.
ASAS improvement criteria, dened as change greater than or
equal to 20% (or an absolute score change of 10; scale 0100) in
three of the AS core domains with no worsening in the fourth
domain, has been recommended to record the efcacy of
SMARD therapy [53]. Relative improvements of 50% (good
improvement) or 70% (dramatic improvement), with an absolute
improvement of 20 (scale 0100), have been proposed follow-
ing treatment with biologics [94]. Improvement criteria have
not been recommended following physical therapy or routine
practice. Smaller score changes following intensive physical
therapy have been reported for the BASDAI and BASFI, with
score change not exceeding 19 and 12 respectively (scale 0100)
[6971]. Several studies reported mean BASFI score change of
less than 7 (scale 0100) [67, 7274]. Moreover, in contrast to
single-item scales, percentage score change in multi-item instru-
ments may be an inappropriate simplication of score change
interpretation, providing erroneous indications of treatment
effectiveness or instrument responsiveness [95]. It is possible
that a 20% score change in the middle of a scale may be quite
different to a 20% change at the ends of the scale [22, 95].
Recent applications of item response theory have shown that a
number of instruments have items that cluster around the middle
of the scale hierarchy [95, 96]. This makes it easier for patients
scoring in the middle of the scale to register score changes
following real changes in health relative to those positioned
nearer the ends of the scale [95]. The level of change in health
that is important to patients, the minimal clinically important
difference (MCID), has not been widely reported and should be
addressed in future research. Instruments should be administered
longitudinally before and after treatment known to improve
HRQL, and health transition questions should be included as
external criteria of change [21].
Although generally less responsive than disease-specic instru-
ments [96], generic instruments have been recommended for
use alongside disease-specic instruments, and their performance
in the assessment of patients with AS should also be considered.
The SF-36 [97] is the most widely used generic health prole
both generally [3] and in studies of AS patients. The SF-36 is the
only generic instrument with evidence of responsiveness following
both drug therapy and physical therapy in AS. The physical
component summary (PCS) score and associated domains
were both responsive [62, 66, 84] and had evidence of discri-
minative validity [62, 65, 66] following anti-TNF therapies.
A concurrent evaluation following anti-TNF therapy reported
comparable levels of responsiveness between the SF-36 PCS
and physical function domain, the BASFI and patients global
assessment of health [66].
In conclusion, this extensive review provides information
necessary for the appropriate selection of AS-specic, multi-
item, patient-assessed health instruments. Such instruments,
specic to the assessment of pain, stiffness, fatigue or global
health, were not identied; where assessed, these domains were
largely measured with single-item visual analogue scales. Single
items may provide a limited reection of these important
domains. These domains are currently inadequately assessed
by AS-specic, multi-item patient-assessed health instruments.
The BASDAI has acceptable measurement properties as a
measure of disease activity, but there are issues relating to item
content and appropriateness of response formats. Although
the BASFI and DFI remain the instruments of choice for
the assessment of functional disability, consensus is required for
583
the recommendation of a single instrument. The domain of
HRQL is important to patients and should also be considered as
a core assessment domain.
Key messages
 Several core assessment domains are
inadequately assessed by AS-specic,
multi-item patient-assessed health instru-
ments. Where assessed, these domains
Rheumatology
are largely measured using single-item
scales.
 Single-item scales may provide a limited
reection of core assessment domains.
 Consensus is required for the assessment
of functional ability.
 Instrument-concurrent evaluation in dif-
ferent settings with long-term follow-up
and support for score interpretation is
required.
The authors have declared no conicts of interest.
Acknowledgements
The authors thank Dr Krysia Dziedzic, Keele University, and
Dr Jonathan Packham, Staffordshire Rheumatology Centre,
for their helpful comments on an earlier version of this review.
We would also like to thank the anonymous reviewers for their
insightful comments.
References
1. Russell AS. Ankylosing spondylitis history. In: Klippel JH,
Dieppe PA, eds. Rheumatology. 2nd edn. London: Mosby, 1998;
1:14.12.
2. Ward MM. Quality of life in patients with ankylosing spondylitis.
Rheum Dis Clin North Am 1998;24:81527.
3. Garratt AM, Schmidt L, Macintosh A, Fitzpatrick R. Quality of life
measurement: bibliographic study of patient assessed health outcome
measures. BMJ 2002;324:141721.
4. Fitzpatrick R, Davey C, Buxton MJ, Jones DR. Evaluating patient-
based outcome measures for use in clinical trials. Health Technol
Assess 1998;2:174.
5. Calin A, Garrett S, Whitelock H et al. A new approach to dening
functional ability in ankylosing spondylitis: the development of
the Bath Ankylosing Spondylitis functional index. J Rheumatol
1994;21:22815.
6. Dougados M, Gueguen A, Nakache JP, Nguyen M, Mery C, Amor B.
Evaluation of a functional index and an articular index in ankylosing
spondylitis. J Rheumatol 1988;15:3027.
7. van der Heijde D, Bellamy N, Calin A, Dougadas M, Khan MA,
van der Linden S. Preliminary core sets for endpoints in ankylosing
spondylitis. J Rheumatol 1997;24:22259.
8. van der Heijde D, Calin A, Dougadas M, Khan MA, van der Linden S,
Bellamy N. Selection of instruments in the core set for DC-ART,
SMARD, physical therapy, and clinical record keeping in ankylosing
spondylitis. Progress report of the ASAS Working Group.
Assessments in Ankylosing Spondylitis. J Rheumatol 1999;26:9514.
9. Streiner DL, Norman GR. Health measurement scales. A practical
guide to their development and use. 3rd edn. Oxford Medical
Publications, 2003.
584 K. L. Haywood et al.
10. Garratt AM, Schmidt L, Fitzpatrick R. Patient-assessed health
outcome measures for diabetes: a structured review. Diabet Med
2002;19:111.
11. McDowell I, Newell C. Measuring health. A guide to rating scales and
questionnaires. 2nd edn. Oxford University Press, 1996.
12. Bowling A. Measuring disease. A review of disease-specic quality of
life measurement scales. Buckingham: Open University Press, 1995.
13. Bowling A. Measuring health. A review of quality of life measurement
scales. Buckingham: Open University Press, 1997.
14. Bakker C, Boers M, van der Linden S. Measures to assess ankylosing
spondylitis: taxonomy, review and recommendations. J Rheumatol
1993;20:172432.
15. Ruof J, Sagha O, Stucki G. Comparative responsiveness of three
functional indices in ankylosing spondylitis. J Rheumatol 1999;
26:195963.
16. Ruof J, Stucki G. Comparison of the Dougados Functional Index
and the Bath Ankylosing Spondylitis Functional Index. A literature
review. J Rheumatol 1999;26:95560.
17. Dagnrud H, Hagen K. Physiotherapy interventions for ankylosing
spondylitis. Cochrane Database Systematic Review 2001;CD002822.
18. Dougados M, Dijkmans B, Khan M, Maksymowych W, van der
Linden S, Brandt J. Conventional treatments for ankylosing
spondylitis. Ann Rheum Dis 2002;61(Suppl. 3):iii4050.
19. Braun J, Sieper J, Breban M et al. Anti-tumour necrosis factor alpha
therapy for ankylosing spondylitis: international experience. Ann
Rheum Dis 2002;61(Suppl. 3):iii5160.
20. Bellamy N. Clinimetric concepts in outcome assessment: the
OMERACT lter. J Rheumatol 1999;26:94850.
21. Beaton DE, Bombardier C, Katz JN, Wright JG. A taxonomy for
responsiveness. J Clin Epidemiol 2001;54:120417.
22. Liang MH, Lew RA, Stucki G, Fortin PR, Daltroy L. Measuring
clinically important changes with patient-oriented questionnaires.
Med Care 2002;40(Suppl.):II4551.
23. Norman GR, Sridhar FG, Guyatt GH, Walter SD. Relation of
distribution and anchor-based approaches in interpretation of changes
in health-related quality of life. Med Care 2001;39:103947.
24. Haywood KL, Hargreaves J, Lamb SE. Multi-item outcome measures
for lateral ligament injury of the ankle: a structured review. J Eval
Clin Pract 2004;10:33952.
25. Haywood KL, Garratt AM, Fitzpatrick R. Quality of life in older
people: a structured review of generic self-assessed health instruments.
Qual Life Res 2004, in press.
26. Lee YS, Schlotzhauer T, Ott SM et al. Skeletal status of men with
early and late ankylosing spondylitis. Am J Med 1997;103:23341.
27. Santos H, Brophy S, Calin A. Exercise in ankylosing spondylitis: how
much is optimum? J Rheumatol 1998;25:215660.
28. Garrett S, Jenkinson T, Kennedy G, Whitelock H, Gaisford P,
Calin A. A new approach to dening disease status in ankylosing
spondylitis: the Bath Ankylosing Spondylitis disease activity index.
J Rheumatol 1994;1:228691.
29. Dziedzic KSG. The Body Chart: a further sketch towards a fuller
picture of ankylosing spondylitis. PhD thesis. University of Keele,
UK, 1997.
30. Jones SD, Steiner A, Garrett SL, Calin A. The Bath Ankylosing
Spondylitis Patient Global Score (BAS-G). Br J Rheumatol
1996;35:6671.
31. Guillemin F, Challier B, Urlacher F, Vancon G, Pourel J. Quality of
life in ankylosing spondylitis: validation of the ankylosing spondylitis
Arthritis Impact Measurement Scales 2, a modied Arthritis Impact
Measurement Scales Questionnaire. Arthritis Care Res 1999;2:15762.
32. Fries JF. Measurement of patient outcome in arthritis. Arthritis
Rheum 1980;23:13745.
33. Daltroy LH, Larson MG, Roberts WN, Liang MH. A modication of
the Health Assessment Questionnaire for the spondyloarthropathies.
J Rheumatol 1990;17:94650.
34. Bakker C, Hidding A, van der Linden S, van Doorslaer E. Cost
effectiveness of group physical therapy compared to individualised
therapy for ankylosing spondylitis. A randomised controlled trial.
J Rheumatol 1994;21:2648.
35. Bell MJ, Bombardier C, Tugwell P. Measurement of functional status,
quality of life, and utility in rheumatoid arthritis. Arthritis Rheum
1990;33:591601.
36. Haywood KL, Garratt AM, Dziedzic K, Dawes PT. Patient centered
assessment of ankylosing spondylitis-specic health related quality
of life: evaluation of the Patient Generated Index. J Rheumatol
2003;30:76473.
37. Nemeth R, Smith F, Elswood J, Calin A. Ankylosing
spondylitis (AS)an approach to measurement of severity and
outcome: Ankylosing Spondylitis Assessment Questionnaire
(ASAQ)a controlled study. Br J Rheumatol 1987;Suppl. 1:6970.
38. Abbott CA, Helliwell, PS, Chamberlain MA. Functional assessment
in ankylosing spondylitisEvaluation of a new self-administered
questionnaire and correlation with anthropometric variables.
Br J Rheumatol 1994;33:10606.
39. Doward LC, Spoorenberg A, Cook SA et al. Development of the
ASQoL: a quality of life instruments specic to ankylosing
spondylitis. Ann Rheum Dis 2003;62:206.
40. Meenan RF, Gertmen PM, Mason JH. Measuring health status in
arthritis. The Arthritis Impact Measurement Scales. Arthritis Rheum
1980;23:14652.
41. Meenan RF, Mason JH, Anderson JJ. AIMS2: the contents and
properties of a revised and expanded Arthritis Impacts Measurement
Scales health status questionnaire. Arthritis Rheum 1992;35:110.
42. Jones SD, Calin A, Steiner A. An update on the Bath Ankylosing
Spondylitis Disease Activity and functional indices (BASDAI,
BASFI): excellent Cronbachs alpha scores [letter]. J Rheumatol
1996;23:407.
43. Haywood KL, Garratt AM, Jordan K, Dziedzic K, Dawes PT.
Disease-specic patient-assessed measures of health outcome in
ankylosing spondylitis: reliability, validity and responsiveness.
Rheumatology 2002;41:1295302.
44. Bostan EE, Borman P, Bodur H, Barca N. Functional disability and
quality of life in patients with ankylosing spondylitis. Rheumatol Int
2003;23:1216.
45. Creemers MCW, vant Hof MA, Fransse MJAM, van de Putte LBA,
Gribnau FWJ, van Riel PLCM. A Dutch version of the Functional
Index for Ankylosing Spondylitis: development and validation in a
long term study. Br J Rheumatol 1994;33:8426.
46. Heikkila S, Viitanen JV, Kautianen H et al. Evaluation of the Finnish
versions of the functional indices BASFI and DFI in spondylarthro-
pathy. Clin Rheumatol 2000;19:4649.
47. Claudepierre P, Sibilia J, Goupille P et al. Evaluation of a
French version of the Bath Ankylosing Spondylitis Disease Activity
Index in patients with spondyloarthropathy. J Rheumatol 1997;
24:19548.
48. Claudepierre P, Sibilia J, Roudot-Thoraval F et al. Factors linked to
disease activity in a French cohort of patients with spondyloarthro-
pathy. J Rheumatol 1998;25:192731.
49. Calin A, Nakache JP, Guegen A, Zeidler H, Mielants H, Dougadas M.
Dening disease activity in ankylosing spondylitis: is a combination of
variables (Bath Ankylosing Spondylitis Disease Activity Index) an
appropriate instrument? Rheumatology 1999;38:87882.
50. Falkenbach A, Franke A, van Tubergen A, van der Linden S.
Assessment of functional ability in younger and older patients
with ankylosing spondylitis: performance of the Bath Ankylosing
Spondylitis Functional Index. Am J Phys Med Rehabil 2002;
81:41620.
51. Auleley GR, Benbouazza K, Spoorenberg A et al. Evaluation of
the smallest detectable difference in outcome or process variables in
ankylosing spondylitis. Arthritis Rheum 2002;47:5827.
52. Eyres S, Tennant A, Kay L, Waxman R, Helliwell P. Measuring
disability in ankylosing spondylitis: comparison of Bath AS
Functional Index with Revised Leeds Disability Questionnaire.
J Rheumatol 2002;29:97986.
53. Anderson JJ, Baron G, van der Heijde D. Ankylosing spondylitis
assessment group preliminary denition of short-term improvement in
ankylosing spondylitis. Arthritis Rheum 2001;44:187686.
 Patient-assessed health in AS: a review
54. Van Tubergen A, Debats I, Ryser L et al. Use of a numerical rating
scale as an answer modality in ankylosing spondylitis-specic
questionnaires. Arthritis Rheum 2002;47:2428.
55. Hidding A, van der Linden SJ, Boers M et al. Is group physical
therapy superior to individualised therapy in ankylosing spondylitis?
Arthritis Care Res 1993;6:11725.
56. Hidding A, van der Linden SJ, Gielen X, de Witte L, Dijkmans B,
Moolenburgh M. Continuation of group physical therapy is necessary
in ankylosing spondylitis: results of a randomised controlled trial.
Arthritis Care Res 1994;7:906.
57. Bakker C, Rutten-van-Molken M, Hidding A, van Doorslaer E,
Bennett K, van der Linden S. Patient utilities in ankylosing spondylitis
and the association with other outcome measures. J Rheumatol
1994;21:1298304.
58. Bakker C, van der Linden SJ, van Santen-Hoeufft M, Bolwijn P,
Hidding A. Problem elicitation to assess patient priorities in ankylosing
spondylitis and bromyalgia. J Rheumatol 1995;22:130410.
59. Calin A, Nakache JP, Gueguen A, Zeidler H, Mielants H,
Dougados M. Outcome variables in ankylosing spondylitis: evaluation
of their relevance and discriminant capacity. J Rheumatol 1999;26:
9759.
60. Braun J, Brandt J, Listing J et al. Treatment of active ankylosing
spondylitis with iniximab: a randomised controlled multicentre trial.
Lancet 2002;359:118793.
61. Temekonidis TI, Alamanos Y, Nikas SN. Iniximab therapy in
patients with ankylosing spondylitis: an open label 12 month study.
Ann Rheum Dis 2003;62:121820.
62. Braun J, Brandt J, Listing J et al. Long-term efcacy and safety of
iniximab in the treatment of ankylosing spondylitis. An open,
observational, extension study of a three-month, randomised,
placebo-controlled trial. Arthritis Rheum 2003;48:222433.
63. Brandt J, Khariouzou A, Listing J et al. Six-month results of a
double-blind, placebo-controlled trial of Etanercept treatment
in patients with active ankylosing spondylitis. Arthritis Rheum
2003;48:166775.
64. Dougados M, Behier JM, Jolchine I et al. Efcacy of celecoxib, a
cyclooxygenase 2-specic inhibitor, in the treatment of ankylosing
spondylitis. Arthritis Rheum 2001;44:1805.
65. Gorman JD, Sack KE, Davis JC Jr. Treatment of ankylosing
spondylitis by inhibition of tumor necrosis factor alpha. N Engl J
Med 2002;346:134956.
66. Wanders AJB, Gorman JD, Davis JC, Landewe RBM, van der
Heijde DMFM. Responsiveness and discriminative capacity of the
assessments in ankylosing spondylitis disease-controlling antirheu-
matic therapy core set and other outcome measures in a trial of
Etanercept in ankylosing spondylitis. Arthritis Rheum 2004;51:18.
67. Heikkila S, Viitanen JV, Kautiainen H, Kauppi M. Does improved
spinal mobility correlate with functional changes in spondyloarthro-
pathy after short term physical therapy? J Rheumatol 2000;27:29424.
68. van Tubergen A, Landewe R, Heuft-Dorenbosch L et al. Assessment
of disability with the World Health Organization Disability
Assessment Schedule II in patients with ankylosing spondylitis. Ann
Rheum Dis 2003;62:1405.
69. Van Tubergen A, Landewe R, van der Heijde D et al. Combined spa-
exercise therapy is effective in patients with ankylosing spondylitis:
a randomized controlled trial. Arthritis Rheum 2001;45:4308.
70. Band DA, Jones SD, Kennedy LG et al. Which patients with
ankylosing spondylitis derive most benet from an in-patient manage-
ment program? J Rheumatol 1997;24:23814.
71. Waldner A, Cronstedt H, Stenstrom CH. The Swedish version of the
Bath Ankylosing Spondylitis Disease Activity Index. Reliability and
validity. Scand J Rheumatol Suppl 1999;111:106.
72. Viitanen JV, Heikkila S. Functional changes in patients with
spondylarthropathy. A controlled trial of the effects of short-term
rehabilitation and 3-year follow-up. Rheumatol Int 2001;20:2114.
73. Sweeney S, Taylor G, Calin A. The effect of a home based exercise
intervention package on outcome in ankylosing spondylitis: a
randomized controlled trial. J Rheumatol 2002;29:7636.
585
74. Heikkila S. The Dougados functional index with the 5-point
Likert scale is sensitive to change due to intensive physiotherapy
and exercise in spondyloarthropathy. Clin Exp Rheumatol 2002;
20:68992.
75. Heikkila S, Viitanen JV, Kautiainen H, Kauppi M. Functional long-
term changes in patients with spondylarthropathy. Clin Rheumatol
2002;21:11922.
76. Cronstedt H, Waldner A, Stenstrom CH. The Swedish version of
the Bath Ankylosing Spondylitis Functional Index. Reliability and
validity. Scand J Rheumatol Suppl 1999;111:19.
77. Van Tubergen A, Boonen A, Landewe R et al. Cost effectiveness
of combined spa-exercise therapy in ankylosing spondylitis: a
randomized controlled trial. Arthritis Rheum 2002;47:45967.
78. Analay Y, Ozcan E, Karan A, Diracoglu D, Aydin R. The
effectiveness of intensive group exercise on patients with ankylosing
spondylitis. Clin Rehabil 2003;17:6316.
79. Stone M, Salonen D, Lax M, Payne U, Lapp V, Inman R.
Clinical and imaging correlates of response to treatment with
iniximab in patients with ankylosing spondylitis. J Rheumatol
2001;28:160514.
80. Ward MM, Kuzis S. Validity and sensitivity to change of spondylitis-
specic measures of functional disability. J Rheumatol 1999;26:1217.
81. Maksymowych WP, Jhangri GS, Leclercq S, Skeith K, Yan A,
Russell AS. An open study of pamidronate in the treatment of
refractory ankylosing spondylitis. J Rheumatol 1998;25:7147.
82. Maksymowych WP, Jhangri GS, Lambert RG et al. Iniximab in
ankylosing spondylitis: a prospective observational inception cohort
analysis of efcacy and safety. J Rheumatol 2002;29:95965.
83. Marzo-Ortega H, McGonagle D, OConnor P, Emery P. Efcacy of
etanercept in the treatment of the entheseal pathology in resistant
spondylarthropathy: a clinical and magnetic resonance imaging study.
Arthritis Rheum 2001;44:21127.
84. Helliwell PS, Marzo-Ortega H, Tennant A. Comparison of a
disease-specic and a generic instrument for measuring health-related
quality of life in ankylosing spondylitis. Arthritis Rheum 2002;
46:3098.
85. Spoorenberg A, van der Heijde D, de Klerk E et al. A comparative
study of the usefulness of the Bath Ankylosing Spondylitis Functional
Index and the Dougados Functional Index in the assessment of
ankylosing spondylitis. J Rheumatol 1999;26:9615.
86. Saag KG. OMERACT 6 brings new perspectives to rheumatology
measurement research. J Rheumatol 2003;30:63941.
87. Dziedzic K. Ankylosing spondylitis. In: David C, Lloyd J, eds.
Rheumatological physiotherapy. London: Mosby, 1998:97114.
88. Bellamy N, Kaloni S, Pope J, Coulter K, Campbell J. Quantitative
rheumatology: a survey of outcome measurement procedures in
routine rheumatology outpatient practice in Canada. J Rheumatol
1998;25:8528.
89. Bellamy N, Muirden KD, Brooks PM, Barraclough D, Tellus MM,
Campbell J. A survey of outcome measurement procedures in
routine rheumatology outpatient practice in Australia. J Rheumatol
1999;26:15939.
90. McHorney CA, Cohen AS. Equating health status measures with
item response theory: illustrations with functional status measures.
Med Care 2000;38(Suppl. II):II-43II-59.
91. Terwee CB, Dekker FW, Wiersinga WM, Prummel MF,
Bossuyt PMM. On assessing responsiveness of health-related quality
of life instruments: guidelines for instrument evaluation. Qual Life Res
2003;12:34962.
92. van Tubergen A, Coenen J, Landewe R et al. Assessment of fatigue
in patients with ankylosing spondylitis: a psychometric analysis.
Arthritis Care Res 2002;47:816.
93. Greenhalgh J, Meadows K. The effectiveness of the use of patient-
based measures of health in routine practice in improving the process
and outcomes of patient care: a literature review. J Clin Eval Clin
Pract 1999;5:40116.
94. Stone MA, Inman RD, Wright JG, Maetzel A. Validation exercise
of the Ankylosing Spondylitis Assessment Study (ASAS) Group
586 K. L. Haywood et al.

response criteria in ankylosing spondylitis patients treated with
biologics. Arthritis Rheum 2004;51:31620.
95. Garratt AM in collaboration with the UK Back Pain Exercise
and Manipulation Trial. Rasch analysis of the Roland Disability
Questionnaire. Spine 2003;28:7984.
96. Stucki G, Daltroy L, Katz JN, Johannesson M, Liang MH.
Interpretation of change scores in ordinal clinical scale and health
status measures: the whole may not equal the sum of the parts. J Clin
Epidemiol 1996;49:7117.
97. Wiebe S, Guyatt G, Weaver B, Matijevic S, Sidwell C. Comparative
responsiveness of generic and specic quality-of-life instruments.
J Clin Epidemiol 2003;56:5260.
98. Ware JE. SF-36 Health Survey manual and interpretation guide.
Boston: Medical Outcomes Trust, 1997.

Clinical Vignette doi:10.1093/rheumatology/keh422

Bilateral calcaneal osteonecrosis in a patient with systemic

lupus erythematosus
Patients with systemic lupus erythematosus (SLE) are at high risk
of the development of osteonecrosis. Common sites of involve-
ment are the femoral head, femoral condyles and proximal
humerus. Calcaneal osteonecrosis is an extremely rare disorder.
The patient is a 41-yr-old man of African origin with SLE
treated with prednisolone, azathioprine and hydroxychloroquine.
In 2003 he presented with pain in his lower legs, maximal in his
heels. On examination there were features of bilateral reex
sympathetic dystrophy and peripheral neuropathy.
Radiographs demonstrated osteopenia in several of the small
bones of the feet. Magnetic resonance imaging showed areas of
osteonecrosis within both calcanei posteriorly; there was also
oedema and atrophy within the lower leg and intrinsic hind foot
muscles, which was symmetrical. A neurophysiological study
conrmed a severe symmetrical, sensorimotor axonal polyneurop-
athy of the lower limbs.
This is the rst clinical case report of a patient with SLE who
developed calcaneal osteonecrosis. Two cases with SLE have been
described previously but in a radiological report that contained
few clinical details [1].
Risk factors in our patient for the development of this rare
complication included treatment with high-dose corticosteroids,
active SLE, peripheral neuropathy and reex sympathetic
dystrophy.

M. W. SEYMOUR, A. W. M. MITCHELL,
C. G. MACKWORTH-YOUNG
1. Abrahim-Zadeh R, Klein RM, Leslie D, Norman A. Characteristics
Correspondence to: M. W. Seymour. E-mail: matthew. of calcaneal bone infarction: an MR imaging investigation. Skeletal
seymour@chelwest.nhs.uk Radiol 1998;27:3214.