close

Вход

Забыли?

вход по аккаунту

?

The relationship of socioeconomic status to subsequent health status in systemic lupus erythematosus.

код для вставкиСкачать
423
BRIEF REPORT
THE RELATIONSHIP OF SOCIOECONOMIC STATUS TO SUBSEQUENT
HEALTH STATUS IN SYSTEMIC LUPUS ERYTHEMATOSUS
JOHN M. ESDAILE, JOHN S. SAMPALIS, DIANE LACAILLE, and DEBORAH DANOFF
We examined the relationship of socioeconomic
status to health status, as determined by the Arthritis
Impact Measurement Scales, in 78 systemic lupus erythematosus patients who had been entered into a prospective study. After controlling for age, disease duration, and disease severity, a significant relationship
between socioeconomic status and outcome was not
demonstrated. All study subjects had health insurance
for medical services. The results have potential implications for health care policy.
In 1979, Ginzler et a1 (1) noted the association
of socioeconomic factors with mortality from systemic
lupus erythematosus (SLE). This aspect of their study
has been criticized (2); nevertheless, the implications
of their results are important. With improved methods
of diagnosis and greater understanding of how to
monitor and treat persons with SLE, the findings of
Ginzler et a1 suggest that further advances in the care
of lupus patients may depend in part on pathogenetic
factors linked to socioeconomic status (SES), or social
forces related to outcome but not necessarily under
the direct control of physicians. Recently, Pincus and
From the Divisions of Rheumatology, Clinical Epidemiology, and Clinical Immunology, Department of Medicine, Montreal
General Hospital and McGill University, Montreal, Quebec, Canada.
Supported in part by a grant from the Lupus Society of
Quebec.
John M. Esdaile, MD, MPH: Divisions of Rheumatology
and Clinical Epidemiology; John S. Sampalis, MSc: Division of
Clinical Epidemiology; Diane Lacaille, BSc: Medical Student; Deborah Danoff, MD: Division of Clinical Immunology.
Address reprint requests to John M. Esdaile, MD, MPH,
Montreal General Hospital, 1650 Cedar Avenue, Montreal, Quebec,
H3G 1A4, Canada.
Submitted for publication April 10, 1987; accepted in revised form August 18, 1987.
Arthritis and Rheumatism, Vol. 31, No. 3 (March 1988)
Callahan (3) have noted the importance of formal
education in determining a variety of outcomes for
persons with rheumatoid arthritis. Furthermore, it has
been suggested that cuts in government funding for
health care in the US may have a major effect on the
health status of the poor and persons with chronic
diseases (4). If socioeconomic factors are associated
with poorer health outcomes in persons with chronic
diseases, there may be implications for health care
funding policies for less advantaged persons with such
disorders.
In Canada, government health plans have provided all residents with hospital care since 1961 and
with nonhospital medical services since 1970. Because
problems of access to health care are thus less likely to
be a cause of adverse outcome, we evaluated the
importance of SES on global health status in a cohort
of persons with SLE who were followed at the Montreal General Hospital.
Patients and methods. All persons meeting the
American Rheumatism Association (ARA) revised criteria for SLE (3,entered into the Montreal General
Hospital Lupus Registry from its inception in January
1977 through December 1983, and followed through
the hospital were considered eligible. Data have been
collected prospectively using standard forms provided
by the ARA Medical Information System.
Following approval of the Montreal General
Hospital Clinical Studies Committee, 85 eligible subjects were identified and followed up in the summer
and fall of 1985. Of the 85 eligible subjects, 5 (6%) had
died; 1 (1%) was known to be alive, well, and receiving
no therapy, but could not be contacted; and I (1%) was
lost to followup. All of the 78 remaining subjects
agreed to be interviewed.
BRIEF REPORTS
424
The zero time for the study was the date of
entry into the lupus registry. Outcome was the total
health status as determined by the self-administered
Arthritis Impact Measurement Scales (AIMS) questionnaire (6), which has been tested for reliability in
SLE (7). Each study subject completed an AIMS
questionnaire and was interviewed to confirm the
education level attained (grade school, high school, or
college) and SES, which was classified using the
British Census Scale (8) as currently used by Statistics
Canada (Joan Lindsay: personal communication).
Both SES and education were determined for the time
of diagnosis, registry entry, and followup in 1985.
The SES scale classifies subjects into 5 categories based on 220 occupations and 4 levels of responsibility (e.g., whether an individual is a supervisor or
foreman). Category 111 is subdivided into nonmanual
(IJIN) and manual (IIIM). Categories and examples
are: I (professionals, e.g., doctors), 11 (managerial and
other professional occupations, e.g., teachers), IIIN
(nonmanual skilled occupations, e.g., salespersons),
HIM (skilled manual workers, e.g., mechanics), IV
Table I .
(partly skilled workers, e.g., orderlies, waitresses),
and V (unskilled workers, e.g., porters, dishwashers).
Data were collected retrospectively from the
registry files for demographic variables (including race
and marital status), duration of SLE from diagnosis to
zero time, and duration of followup. Data were also
extracted for the pre-zero time interval, zero time, and
the post-zero time interval for treatment (nonsteroidal
antiinflammatory drugs, prednisone, immunosuppressive agents) and for 2 measures of disease severity.
The first measure was the ARA lupus criteria count, as
used by Ginzler et a1 (1). The second, the Lupus
Activity Criteria Count (LACC) (9), was used as a
scale of 0-7. The latter was selected because it resembled the ARA criteria count but assigned a lower
weight for dermatologic involvement. Zero time data
were obtained for potentially important laboratory
predictors (hematocrit, leukocyte count, dipstick findings for urinary protein, and serum creatinine, albumin, C3, C4, and DNA binding). Post-zero time data
were extracted on the number of hospitalizations and
outpatient visits.
Characteristics of study population at entry to registry
No. (5%)
Vanable*
Female
Race
White
Black
Oriental
Marital status
Single
Married/living as married
Divorced/widowed
Education
College
Completed
Not completed
High school
Completed
Not completed
Grade school
Completed
Not completed
SES
I
I1
111 nonmanual
111 manual
IV
V
* SES
Total
(n = 78)
Deceased
(n = 5)
Not contacted
(n = 2)t
73 (94)
2 (100)
67 (86)
1 (50)
1 (50)
0
8 (10)
3 (4)
23 (29)
48 (62)
7 (9)
0
0
2 (100)
7 (9)
0
0
14 (18)
20 (26)
0
15 (19)
8 (10)
0
0
10 (13)
20 (26)
25 (32)
14 (18)
9(II)
0
0
0
1
0
0
0
14 (18)
1
= socioeconomic status, classified into categories I-V, using the British Census Scale. See
Patients and Methods for details.
t Data on education and SES were not available for 1 subject.
BRIEF REPORTS
425
Table 2. Relationships of potential predictors (at zero time) to health status, as determined by
AIMS questionnaire. in 78 patients with SLE*
Variable?
Mean
SES
Education
Age (years)
Followup after entry (years)
SLE duration before entry (years)
LACC
ARA criteria
Hematocrit (%)
Platelets ( x lo3)
WBC ( x 10’)
Serum albumin (mg/dl)
Creatinine (mg/dl)
DNA binding (%)
C3 (mg/dl)
C4 (mg/dl)
Urinary protein (0-4+)
-
42
5.9
4.4
2.3
5.6
37
267
6.0
3.9
0.8
37
I23
18
0.4
Range
17-25
2-9
0-25
0-5
4-9
14-47
68-625
0.7-1 5.5
2.1-5.4
0.4-2.2
0-98
26-250
1-85
0-3
Pearson
correlation
coefficient
0.14
0.19
0.35
0.26
0.22
0.17
0.01
- 0.03
0.01
0.03
0.04
- 0.05
0.06
- 0.04
0.07
- 0. I9
P
0.22
0.10
<0.01
<0.05
0.05
0. I4
0.97
0.80
0.90
0.80
0.76
0.69
0.59
0.74
0.56
0.12
* AIMS = Arthritis Impact Measurement Scales (6); SES = socioeconomic status; - = not done;
SLE = systemic lupus erythematosus; LACC = Lupus Activity Criteria Count (9); ARA criteria =
American Rheumatism Association criteria ( 5 ) ; WBC = white blood cells.
t SES was classified into categories I-V (8),using the British Census Scale; education was scored on
a scale of 1-6 (see Patients and Methods for details). Zero time data on the platelet count were missing
for 3 subjects, and zero time data on urinary protein were missing for 7 subjects.
Multiple linear regression (SAS, 1985)was used
to generate an explanatory model of the variation in
total health status. The potential independent variables included the demographic, temporal, severity,
and laboratory variables. Model selection was based
on examination of R2, residual mean square, and
Mallows’ Cp statistic (10). The SES and education
variables were then added separately to this basic
model to test their ability to provide additional predictive information. SES and education were evaluated as
linear (continuous) variables and as class (dummy)
variables. Residual analysis and normal probability
plots were used to evaluate the appropriateness of
multiple linear regression for the data set (10).
Results. The characteristics of the study population and the correlation of selected variables to the
total health score derived from the AIMS questionnaire are described in Tables 1 and 2. In that the
outcome was defined as total health status (determined
by the AIMS questionnaire), the 5 deceased subjects
were not included in the analysis; however, none of
those patients were in SES category I, and all of the
patients in SES category V died (Table 1). Using the
AIMS, the mean + 1 SD of the total health score was
1.94 + 1.67 for the 78 patients contacted.
The multivariate model selected included age at
registry entry, the LACC at entry, and the duration of
SLE from diagnosis to entry (Table 3). Stepwise
forward and backward multiple linear regression demonstrated the identical model. Duration of followup
from zero time to questionnaire administration had
been hypothesized a priori to be an important covariate, and this variable was forced into the basic model.
The effect of all two-way interactions was examined,
and none were found to be significant.
SES and formal education were correlated (r =
0.36, P = 0.001). The additional effect of SES and
education was tested by adding these variables to the
model separately, and the results are summarized in
Table 3. Education did not add significantly to the
variance explained when added to the multivariate
model as a linear or a class variable. There was a trend
suggesting an effect of SES as a linear variable, and as
a class variable defined as nonmanual (I, 11, IIIN) or
manual (IIIM, IV); no subject was classified into
category V. SES classified into terciles of high (I, II),
middle (IIIN), and low (IIIM, IV) demonstrated that
those in the highest tercile had a significantly better
health status (lower AIMS score) than those in the
lowest tercile. When the multiple hypotheses tested
were taken into account, however, no SES variable
had a significant ef€ect.
BRIEF REPORTS
426
There were no significant interactions between
SES as a linear or class variable and the other variables included in the multivariate model. There was no
association with either SES or total health status for
any of the measures of treatment or the number of
post-zero time outpatient visits or hospitalizations,
suggesting that these effects could not explain the
results. Changes in SES from diagnosis to study entry
or from entry to followup were assessed and did not
explain the observed results.
To further evaluate the relationship between
SES and total health score, the latter was dissected into
3 dimensions: Physical, Depression, and Pain. The
Physical dimension consisted of the average of the
standardized results for the Mobility, Dexterity, Physical Activity, and Household Activity subscales of the
total health score. Depression and Pain are individual
subscales of the AIMS instrument. For the bivariate
comparisons of SES rated 1-5 with the Physical, Depression, and Pain components, the Pearson correlation
coefficients were 0.07 ( P > O S O ) , 0.13 (P > 0.25), and
0.28 (P < 0.02), respectively. When SES as a class
variable (manualhonmanual or high/middleAow) was
contrasted with the Physical, Depression, and Pain
measures, and when stepwise multivariate analyses
with SES were conducted with these same outcomes,
SES was associated only with pain (data not shown).
Discussion. After controlling for age, disease
severity, and disease duration before and after study
entry, the relationship of SES and education to subsequent health status in SLE was studied (Table 3). The
level of formal education was related to health status
in a bivariate analysis (Table 2 ) , but no independent
effect of education was present once the covariates
were taken into account (Table 3). There was a trend
toward an effect of SES on global health status,
particularly when comparing the AIMS scores of the
highest versus the lowest socioeconomic groups
(Table 3), but the effect was not statistically significant
when the multiple hypotheses tested were considered.
A possible explanation for our results is that the
AIMS is not an appropriate outcome instrument for
SLE. While no instrument has been specifically validated for use in SLE, the AIMS has been validated in
rheumatoid arthritis and osteoarthritis (7), has undergone reliability testing in SLE (7), and has considerable
content validity for SLE. Furthermore, Liang et al ( 1 1)
have noted that the major outcomes in SLE are the loss
of physical function and independence and the development of depression. For this reason, we conducted a
secondary analysis to specifically evaluate the relationship of SES to physical ability and depression. No
association was found. This suggests that SES was not
related to outcomes considered important in SLE.
Table 3. Results of multivariate analysis
Standard
error
P value for
independent
effect of
variable
Model
R2
0.04
0.35
0.05
0. I9
0.01
0.14
0.03
0.09
0.09
0.267
0.04
Effect of SES variables
added to basic model
Continuous variable (1-5)
Manual versus nonmanual
I , I1 vs. HIM, IV
I , I1 vs. IIIN
IIIN vs. IIIM, IV
0.26
0.64
- 0.95
- 0.68
- 0.26
0.13
0.36
0.39
0.38
0.40
0.06
0.08
Effect of education variables
added to basic model
Continuous variable (1-6)
College vs. grade school
College vs. high school
High school vs. grade school
0.09
-0.64
- 0.04
-0.60
0.11
0.45
0.40
0.40
0.42
Variable*
Coefficient
Basic model
Age (years)
LACC (0-7)
Duration of SLE before entry (years)
Followup after entry (years)
* See Table 2 for definitions and explanations.
0.303
0.298
0.329
0.52
0.274
0.294
0.14
BRIEF REPORTS
Ginzler et a1 (1) assessed the effect of SES on
mortality in 1,013 patients with SLE (mean followup
44 months) by contrasting mortality rates of persons
who received public funding for health care with
mortality rates of those with private funding, including
Medicare. They found a significant association between low SES (public funding) and higher mortality at
1 , 5 , and 10 years after study entry.
Important differences exist between the present
study and that of Ginzler et al. First, Ginzler’s patients
probably had more severe disease. Only 2 of our study
subjects (2%) had azotemia (as defined by Ginzler) at
entry, in contrast to 27% of patients in the Ginzler
study. Ginzler’s patients had a 5-year survival rate
from entry of 77%; in our subjects, the 5-year survival
rate was 96% (calculated in the same manner). Second, Ginzler et a1 evaluated mortality, whereas we
examined health status using the total health score
from the AIMS. We used the AIMS score not only
because of the comparatively low mortality rate and
the small sample size available, but also because with
increased survival in SLE, intermediate outcomes
such as health status are increasingly relevant to both
physicians and patients. Third, financial constraints
are less likely to explain the results in our patients,
because a universal medical insurance scheme had
been in place for 7 years prior to the enrollment of the
first patient into the registry. In contrast, 57% of the
patients in the study by Ginzler et a1 lacked private
insurance o r Medicare.
Social factors as determinants of outcome in
patients with rheumatic diseases have recently received increased attention (1,3,12,13). Health education may influence outcome in persons with arthritis
(14), and the ability to avail oneself of such education
o r to absorb or employ the recommendations may vary
among social classes. Markers of social class are likely
surrogates for other explanatory mechanisms (3), such
as differences in access to health care.
The increased access to health care providers
permitted by a universal health insurance scheme is a
possible explanation for the absence of an association
between SES and global health status in our patients
with SLE. Inequities in access may have a major effect
on the health status of patients with chronic diseases
(4); therefore, the relationships between social factors,
access to health care, and outcome clearly merit
further study in SLE and other chronic rheumatic
diseases.
427
Acknowledgments. The authors acknowledge the assistance of Joan Lindsay in the classification of socioeconomic status and of Samy Suissa, PhD, in the data analysis,
and the secretarial expertise of Claire Boudreau.
REFERENCES
1. Ginzler EM, Diamond HS, Weiner M, Schlesinger M,
Fries JF, Wasner C, Medsger TA Jr, Ziegler G, Klippel
JH, Hadler NM, Albert DA, Hess EV, Spencer-Green
G, Grayzel A, Worth D, Hahn BH, Barnett EV: A
multicenter study of outcome in systemic lupus erythematosus. I. Entry variables as predictors of prognosis.
Arthritis Rheum 25:601-611, 1982
2. Larson MG, Liang MH: The multicenter study of outcome in systemic lupus erythematosus:a critique (letter).
Arthritis Rheum 26570-571, 1983
3. Pincus T, Callahan LF: Formal education as a marker
for increased mortality and morbidity in rheumatoid
arthritis. J Chronic Dis 38:973-984, 1985
4. Mundinger MO: Health service funding cuts and the
declining health of the poor. N Engl J Med 313:44-47,
1985
5. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ,
Rothfield NF, Schaller JG, Tala1 N, Winchester RJ: The
1982 revised criteria for the classificationof systemic lupus
erythematosus. Arthritis Rheum 25:1271-1277, 1982
6. Meenan RF, Gertman PM, Mason JH: Measuring health
status in arthritis: the Arthritis Impact Measurement
Scales. Arthritis Rheum 23:146-152, 1980
7. Meenan RF, Gertman PM, Mason JH, Dunaif R: The
Arthritis Impact Measurement Scales: further investigation of a health status measure. Arthritis Rheum 25:
1048-1053, 1982
8. Leete R, Fox J: Registrar General’s social classes:
origins and uses. Popul Trends 8: 1-7, 1977
9. Urowitz MB, Gladman DD, Tozman ECS, Goldsmith
CH: The lupus activity criteria count. J Rheumatol
1 1 1783-787, 1984
10. Montgomery DC, Peck EA: Introduction to Linear Regression Analysis. New York, John Wiley & Sons, 1982
1 1 . Liang MH, Rogers M, Larson M, Eaton HM, Murawski
BJ, Taylor JE, Swafford J, Schur PH: The psychosocial
impact of systemic lupus erythematosus and rheumatoid
arthritis. Arthritis Rheum 27:13-19, 1984
12. Yelin E, Meenan R, Nevitt M, Epstein W: Work disability in rheumatoid arthritis: effects of disease, social, and
work factors. Ann Intern Med 9331-556, 1980
13. Cunningham LA, Kelsey JL: Epidemiology of musculoskeletal impairments and associated disability. Am J
Public Health 74574-579, 1984
14. Lorig K, Lubeck D, Kraines RG, Seleznick M, Holman
HR: Outcomes of self-help education for patients with
arthritis. Arthritis Rheum 28:680-685, 1985
Документ
Категория
Без категории
Просмотров
0
Размер файла
431 Кб
Теги
statue, lupus, health, subsequent, systemic, erythematosus, relationships, socioeconomic
1/--страниц
Пожаловаться на содержимое документа