close

Вход

Забыли?

вход по аккаунту

?

Evaluation of the european spondylarthropathy study group preliminary classification criteria in alaskan eskimo populations.

код для вставкиСкачать
534
EVALUATION OF THE
EUROPEAN SPONDYLARTHROPATHY STUDY GROUP
PRELIMINARY CLASSIFICATION CRITERIA IN
ALASKAN ESKIMO POPULATIONS
GEORGIANA S. BOYER, DAVID W. TEMPLIN, and WENDELL P. GORING
Objective. To evaluate the preliminary classification criteria proposed by the European Spondylarthropathy Study Group (ESSG) in Alaskan Eskimo populations.
Methods. We examined, interviewed, and reviewed the records of 104 Eskimo patients with spondylarthropathy and 75 with other rheumatic disorders, and
evaluated them according to the proposed criteria.
Results. We found an overall sensitivity of 88.5%
and a specificity of 89.3%, which is similar to the
reported values in European populations.
Conclusion. The ESSG criteria performed well in
a population very different from that in which they were
developed, and deserve further evaluation as a muchneeded and useful epidemiologic tool.
Epidemiologic studies of the diseases which
make up the spondylarthropathies (SpA) have been
hampered by the lack of adequate disease criteria
(1-4). In earlier studies of rheumatic disease prevalence in native populations of Alaska, we found that
_ ~ ~ _ _ _ _
The opinions and assertions contained herein are those of
the authors and do not necessarily represent those of the Indian
Health Service.
From the Alaska Area Native Health Service, Anchorage,
and the University of Arizona, Tucson.
Supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases and by Indian Health Service Interagency Agreement Y02-AR-00005-02.
Georgiana S . Boyer, MD: Consultant in Epidemiology,
Alaska Area Native Health Service; David W. Templin, MD:
Consultant in Rheumatology, Alaska Area Native Health Service;
Wendell P. Goring, MS: Coordinator of Computer Activities, Office
of the Vice President for Research, University of Arizona.
Address reprint requests to Georgiana S . Boyer, MD, 500
East Rudasill Road, Tucson, AZ 85704.
Submitted for publication June 9, 1992; accepted in revised
form November 19, 1992.
Arthritis and Rheumatism, Vol. 36, No. 4 (April 1993)
one-third of the cases of probable SpA could not be
classified according to existing criteria for ankylosing
spondylitis (AS), Reiter’s syndrome (RS) and other
reactive arthritides (ReA), inflammatory bowel disease
(IBDkrelated arthritis, or psoriatic arthritis (PsA).
Rather than exclude the unclassifiable cases from
consideration, we developed working criteria for undifferentiated spondylarthropathy (USpA), based on
the features observed in most of our unclassifiable
cases (3,4). Recently, the European Spondylarthropathy Study Group (ESSG) developed classification
criteria intended to encompass a broader spectrum of
spondylarthropathic disease, with the specific intent of
including previously neglected cases of undifferentiated disease (5). We have tested these criteria in a
wide variety of rheumatic disorders in Alaskan Eskimos, in order to contribute to the necessary validation
of these preliminary criteria. We report the results of
our evaluation herein.
PATIENTS AND METHODS
A total of 104 patients with SPA were included in the
evaluation. Sixty-six patients were identified during ongoing
epidemiologic studies in the Barrow and Bristol Bay Health
Service Units of the Alaska Area Native Health Service
(AANHS), and 38 patients were known cases from previous
epidemiologic studies in Kotzebue and the YukonKuskowim Delta (3,4). Initial identification of possible cases
in the Barrow and Bristol Bay areas was accomplished
primarily through the computerized patient care information
system used by the AANHS (6,7). The medical records of
patients identified as possible cases were then reviewed by
the investigators. Those patients considered likely to have
SpA were invited to attend special clinics for interview and
examination in order to verify their diagnoses and collect the
data necessary for evaluating the ESSG criteria ( 5 ) .
EVALUATION OF ESSG CRITERIA IN ALASKAN ESKIMOS
All 104 patients included in this study fulfilled one of
the following sets of criteria: for AS, the New York criteria
(8); for RS, the 1982 preliminary criteria of the American
College of Rheumatology (ACR) (formerly, the American
Rheumatism Association) (9); for IBD-related arthritis, PsA,
and ReA following dysenteric illness, the diagnostic guidelines specified in the ninth edition of the Primer on the
Rheumatic Diseases (lo), or for USpA, our previously
developed working criteria (3,4). These working criteria
consisted of objective evidence of inflammatory joint involvement observed by a physician plus any 3 items from the
following list: as part of the presenting illness, asymmetric
joint involvement, predominantly of the lower extremities;
enthesopathy; sterilejoint effusion; persistent low back pain
or physical findings consistent with sacroiliitis; mild anemia
and elevated erythrocyte sedimentation rate, not accounted
for by the presence of other disease; or, based on historical
information in the patient’s medical record, evidence of an
intermittent disease course, with recurrent episodes of joint
involvement or the occurrence of iritis or uveitis, diagnosed
by a physician. Exclusions consisted of the disorders specified in the ACR’s 1958 revised criteria for rheumatoid
arthritis (RA) (1 1).
We recruited 75 control subjects from consecutive
patients being seen for other rheumatic diseases at the same
rheumatology clinics as the SpA patients in the towns of
Barrow, Kotzebue, Bethel, Dillingham, and Anchorage, as
well as in the villages surrounding those towns.
All study subjects were of Alaskan Eskimo heritage.
All were examined by the same rheumatologist (DWT), and
were interviewed according to a standardized protocol that
covered all historical information required by the ESSG
criteria, including the features of inflammatory back pain.
Their records were also reviewed thoroughly. Throughout
the study we adhered to the definitions specified by the
ESSG for each of the variables (5). Standard anteroposterior
radiographs of the pelvis, for examination of the sacroiliac
(SI)joints, were performed only when clinically indicated, in
accordance with our human research committee guidelines.
Radiographs permitting adequate visualization of the SI
joints were obtained for 91 of the 104 SpA patients, but for
only 47 of the 75 control patients. Radiographs of study
patients were admixed with those of non-study subjects, and
were read by the study rheumatologist (DWT), who had no
knowledge of the patients’ clinical diagnoses. Changes were
graded and scored as specified in the Population Studies of
the Rheumatic Diseases guidelines (12).
All clinical, demographic, and radiographic data
were entered in a commercially available database program
(Paradox) running on a microcomputer. Using modular
scripts, the individual subjects were scored for diagnostic
points of the ESSG criteria. The ESSG criteria specify that
a subject must have either inflammatory spinal pain or
synovitis, asymmetric or predominantly in the lower limbs,
plus one other feature. For scoring purposes in our study,
either inflammatory spinal pain or asymmetric synovitis was
given 5 points. Only if inflammatory spinal pain or asymmetric synovitis was present was an additional point awarded for
the presence of any one of the following features: positive
family history of SPA; psoriasis; IBD; alternate buttock
pain; enthesopathy, specifically involving the plantar fascia
535
or Achilles tendon; nongonococcal urethritis or cervicitis or
acute diarrhea occurring within 1 month before arthritis; or
sacroiliitis, bilateral grade 2 2 or unilateral grade 2 3 . A
minimum of 6 points was required to fulfill the criteria.
Modifications of the criteria were tested by making selected
changes in the modular scripts.
RESULTS
Patient characteristics. The demographic characteristics of the 104 Eskimo patients with SPA and
the 75 control subjects with other rheumatic disorders,
their diagnoses, and the mean number of years since
the onset of first symptoms are summarized in Table 1.
No significant demographic differences between cases
and controls were observed. Although no effort was
made to choose subjects to match those in the ESSG
study, the average age, ratio of men to women, and
duration of disease among the Eskimo patients were
all quite similar to those of the European subjects (5).
The relative frequencies of specific types of
SPA, however, were quite different in the Eskimo and
European populations. In our study, ReA, which was
nearly always associated with nonspecific urethritis or
cervicitis, was the most frequent type of SPA, almost
twice as common as AS. USpA was second in frequency, accounting for approximately one-third of the
cases, and PsA was quite rare. In the European group,
AS was the most common condition, followed in order
by USpA, PsA, and ReA.
Performance of the ESSG criteria in the Eskimo
population. Ninety-two of the 104 cases of SpA met
the ESSG criteria, yielding an overall sensitivity of
88.5%. Among the 12 cases that failed to fulfill the
criteria, 8 had USpA and 4 had RS. The sensitivities of
the criteria in different subgroups of SpA in the Eskimo and ESSG studies are compared in Table 2. In
our study, sensitivity ranged from a low of 75% in
USpA to a high of 100% in AS. The ESSG reported
sensitivities varying from 78.4% for USpA to 93.6%
for AS and 100% for IBD-related arthritis.
Of the 75 Eskimo control patients with other
rheumatic diseases, 8 fulfilled the ESSG criteria; this
yielded an overall specificity of 89.3%, compared with
87% reported by the ESSG. The Eskimo control
subjects who met the criteria included 3 with RA, by
both the 1958 (11) and the 1987 (13) ACR criteria; 3
with posttraumatic degenerative arthritis and back
pain; 1 with healed tuberculosis of the SI joint; and 1
with previous tuberculosis of the knee and ankle.
The sensitivities and specificities for each of the
individual criteria of the ESSG in Eskimo and Euro-
536
BOYER ET AL
Table 1. Selected demographic features of 104 Eskimo patients with spondylarthropathy and 75
control patients with other rheumatic diseases*
~
~
~~~
No. of
Age
Disease duration
patients (mean f SD) % male (mean 2 SD years)
Disease
Spondylarthropathy group
Reactive arthritis (including RS)
Undifferentiated SpA
Ankylosing spondylitis
Other SpA (IBD, PsA)
All patients with SpA
Control group with other rheumatic diseases
Rheumatoid arthritis
Osteoarthritis
Trauma-related back and/or joint pain
Mechanical back pain, tendinitis,
entrapment syndromes
Other rheumatic disease (SLE, gout, septic
or tuberculous arthritis)
Miscellaneous soft tissue disease and
arthralgid
All control patients with other rheumatic
diseases
45
32
25
2
104
43.2 f 13.5
43.3 t 14.4
48.9 f 12.76
49t
44.7 t 13.8
76
42
67
50
64
17.0 -C 11.0
16.2 f 10.4
24.7 t 8.9
12.5t
18.5 f 11.0
23
16
16
10
51.5 5 15.9
57.4 f 5.2
36 f 12.5
45.4 -+ 16.1
22
31
75
50
11.8 t 8.6
12.6 2 7.2
10 2 4.7
16.3 f 11.1
6
58.37
33
30.7t
4
56t
0
15.8t
75
49.4 5 15.5
39
13.8
f
11.1
* RS = Reiter’s syndrome; SpA = spondylarthropathy; IBD = inflammatory bowel disease; PsA
psoriatic arthritis; SLE = systemic lupus ervthematosus.
t Sample too small for standard deviation.
pean subjects are compared in Table 3. The higher
frequency of ReA in the Eskimo population is reflected
in the high sensitivity figures for asymmetric synovitis
(82.7%) and infection preceding arthritis (56%), and in
the lower sensitivity of inflammatory back pain (46%),
a reversal of the European experience (Table 3).
However, for the criteria such as alternate buttock
pain and enthesopathy, which are not specifically tied
to a particular syndrome, the sensitivity and specificity
figures in Eskimo and European populations are remarkably similar.
Of the 91 SpA patients who had pelvic radiographs, 49 had bilateral sacroiliitis of grade 2 2 or
unilateral involvement of grade 8 3 , yielding a sensi-
Table 2. Sensitivity of the ESSG criteria in subgroups of the
spondylarthropathies in Eskimo and European patients*
Eskimo
Disease subgroup
Ankylosing spondylitis
RSlreactive arthritis
Undifferentiated SpA
Psoriatic arthritis
IBD-related arthritis
European
No. of Sensitivity No. of Sensitivity
patients
(%)
patients
(%)
25
45
32
1
1
100
88
75
100
I00
157
36
102
49
17
93.6
80.6
78.4
81.6
100
* See reference 5 for complete data on the European Spondylarthropathy Study Group (ESSG) patients; see Table 1 for definitions.
=
tivity of 54%. Because 13 patients did not have pelvic
radiographs, we can only estimate the percentage of
total cases with sacroiliitis to be between 47% (the
Table 3. Sensitivity and specificity of the 9 variables of the ESSG
criteria in Eskimo and European patients*
Sensitivity (%)
Variable
Asymmetric
synovitis
Inflammatory back
pain
Positive family
history
Urethritis, cervicitis,
acute diarrhea
within 1 month
before arthritis
Alternate buttock
pain
Enthesopathy
(Achilles tendon,
plantar fascia)
Psoriasis
Inflammatory bowel
disease
Radiographic
sacroiliitist
Specificity (%)
Eskimos Europeans Eskimos Europeans
82.7
41.3
91.7
87.3
46
74.6
83.6
82.5
31
32.2
94.5
94.5
56
19.1
93
97
20
20.4
89
97.3
25
36.5
90.4
88.9
22.7
9.6
95.9
98.6
95.2
97.3
54.4
89.4
97.8
8.7
1
54
* See reference 5 for complete data on the European Spondylarthropathy Study Group (ESSG) patients.
t Represents 54% of the 91 Eskimo cases and 89.4% of the 75
Eskimo controls with radiographic data available.
EVALUATION OF ESSG CRITERIA IN ALASKAN ESKIMOS
expected rate if none of the 13 had sacroiliitis) and 60%
(if all had sacroiliitis). Because the patients without
symptoms referable to the SI joints were less likely to
have undergone radiographic examination, and are
presumably less likely to have sacroiliitis, we believe
the true prevalence of sacroiliitis in the entire SpA
patient group is closer to 47% than to 60%.
Films of the SI joints were available for only
47 of 75 control patients. Five had radiographic sacroiliitis, a specificity of 89.4%. One control subject
had grade 3 unilateral sacroiliac disease; the remaining 4 had minimal grade 2 bilateral sclerosis. Two of
the latter 4 had clear-cut RA (by the 1987 criteria of
the ACR).
Performance of the ESSG criteria in the absence
of radiographic sacroiliitis. Because criteria that do not
involve a need for pelvic radiographs would be extremely useful, especially for epidemiologic studies,
we tested alternate criteria and scoring systems in our
data set. When we tested the outcome of not awarding
a point for radiographic sacroiliitis, the overall sensitivity decreased to 78%. When radiographic evidence
of sacroiliitis was dropped as a feature, 7 patients with
AS, 3 with USpA, and 1 with RS fell short of the 6
points needed to fulfill the criteria. We also tested a
modification that did not include radiographic data but
awarded an extra point when a patient had asymmetric
or lower limb synovitis and inflammatory back pain,
making it possible to meet the threshold score of 6
without the presence of any other variables. In effect,
this modification gave the presence of the second
“major” criterion the same weight in the scoring as
any of the 7 “minor” variables specified in the criteria,
such as enthesopathy, psoriasis, or positive family
history. Using this modification we found a sensitivity
of 82%, with no decrease in specificity. Two patients
with AS, 1 with RS, and 1 with USpA had both
asymmetric synovitis and inflammatory back pain; the
presence of radiographic sacroiliitis was not needed to
score the 6 points needed to fulfill the criteria.
DISCUSSION
Our evaluation of the preliminary ESSG criteria
for spondylarthropathy in Alaskan Eskimos yielded an
overall sensitivity of 88.5% and a specificity of 89.3%.
The ESSG values were 86.7% and 87%, respectively.
It is encouraging to find such similar performance of
the criteria in the Eskimo and European populations,
despite the very different cultural and geographic
537
settings, and despite differences in individual disease
frequencies.
The ESSG criteria were developed with the
intention of encompassing a broader spectrum of disease than previous criteria, with emphasis on including
the previously unclassifiable undifferentiated spondylarthropathies. As in the European study, we found
that the criteria included approximately 75% of the
undifferentiated cases. Although 75% clearly is far
better than zero, there remains room for improvement
in the criteria for undifferentiated disease.
It is important to stress that neither we nor
others are proposing that USpA constitutes a new and
distinct disease entity (2,4,14). The term provides a
useful label for otherwise unclassifiable syndromes
that appear to be part of the SpA spectrum and which
need further study; to recognize their existence is
necessary to resolving the nosologic problems that
have hindered previous epidemiologic studies of these
disorders.
For the ESSG to have widespread application
as classification criteria, the questions required to
meet the criteria must be asked routinely by clinical
providers attending to patients with back pain and
synovitis, and the responses must appear in the patients’ records. Even with the advantage of patient
interview and examination, we often relied on the
medical records to document the occurrence of clinical
features included in the criteria. With many patients
we found considerable difficulty in estimating the
duration of back pain, to such an extent that the time
element rarely contributed to the 4 of 5 features
necessary for determining inflammatory back pain.
Part of the problem may lie in the different sense of
time and priorities concerning symptoms in the Eskimo patients. Quite understandably, they were more
interested in discussing current symptoms than details
of past complaints.
In general, however, the successful performance of the ESSG criteria in the Alaskan setting
bodes well for their usefulness as classification criteria. Worth emphasizing is the rather good performance
of the ESSG criteria even when pelvic radiographs are
not available and radiographic sacroiliitis is not included as a criterion. We propose a minor modification
of the criteria which, in our study population, served
to increase the sensitivity somewhat when radiographic sacroiliitis was not included among the criteria. We hope that other investigators will also test
modifications that may minimize the need for pelvic
radiographs in future epidemiologic studies.
BOYER ET AL
538
REFERENCES
1. Khan MA, van der Linden SM: A wider spectrum of
2.
3.
4.
5.
6.
7.
spondyloarthropathies. Semin Arthritis Rheum 20: 107113, 1990
Thomson GTD, Inman RD: Diagnostic criteria in the
spondyloarthropathies: towards a base for revised nosology. J Rheumatol 17:426-429, 1990
Boyer GS, Lanier AP, Templin DW: Prevalence rates of
spondyloarthropathies, rheumatoid arthritis and other
rheumatic disorders in an Alaskan Inupiat Eskimo population. J Rheumatol 15:678-683, 1988
Boyer GS, Lanier AP, Templin DW, Bulkow L:
Spondyloarthropathy and rheumatoid arthritis in Alaskan Yupik Eskimos. J Rheumatol 17:489-496, 1990
Dougados M, van der Linden S, Juhlin R, Huitfeldt B,
Amor B, Calin A, Cats A, Dijkmans B, Olivieri I, Pasero
G. Veys E , Zeidler H: The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 34: 12181227, 1991
Brown GA: Patient Care Information System: a description of its utilization in Alaska, The Fourth Annual
Symposium on Computer Applications in Medical Care,
Washington, DC, 1980
Mason W: Patient Care Component (PCC) Overview.
Tucson, AZ, Office of Health Program Research and
Development, 1990
8. Moll JMH, Wright V: New York clinical criteria for
ankylosing spondylitis: a statistical evaluation. Ann
Rheum Dis 32:354-363, 1973
9. Willkens RF, Arnett FC, Bitter T, Calin A, Fisher L,
Ford DK, Good AE, Masi AT: Reiter’s syndrome:
evaluation of preliminary criteria for definite disease.
Arthritis Rheum 24:844-849, 1981
10. Schumacher HR Jr, Klippel JH, Robinson DR, editors:
Primer on the Rheumatic Diseases. Ninth edition. Atlanta, Arthritis Foundation, 1988
1 1 . Ropes MW, Bennett GA, Cobb S, Jacox R, Jessar RA:
1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9:175-176, 1958
12. Bennett PH, Wood PHN, editors: Population Studies of
the Rheumatic Diseases. Amsterdam, Excerpta Medica,
1968
13. Arnett FC, Edworthy SM, Bloch DA, McShane DJ,
Fries JF, Cooper NS, Healey LA, Kaplan SR, Liang
MH, Luthra HS, Medsger TA Jr, Mitchell DM, Neustadt DH, Pinals RS, Schaller JG, Sharp JT, Wilder RL,
Hunder GG: The American Rheumatism Association
1987 revised criteria for the classification of rheumatoid
arthritis. Arthritis Rheum 31:315-324, 1988
14. Zeidler H, Mau W, Khan MA: Undifferentiated spondyloarthropathies. Rheum Dis Clin North Am 18: 187-202,
1992
Документ
Категория
Без категории
Просмотров
0
Размер файла
458 Кб
Теги
population, spondylarthropathy, group, stud, evaluation, alaska, classification, criterias, eskimo, european, preliminary
1/--страниц
Пожаловаться на содержимое документа