- Top of page
- SUBJECTS AND METHODS
The role of arthritis patients as informed participants in their health care has become increasingly important. Patients with higher education levels tend to be more proactive in their health care and have better outcomes (1). To improve patient education in arthritis, several programs were developed, including the Stanford Arthritis Self-Help Course (ASHC) in 1979, which produced improvements in self efficacy and pain reduction (2) that were sustained at 4 years (3), and helped reduce health care costs and hospitalization (3–5). The cornerstone of the self-management program was found to be the self-efficacy component rather than knowledge or behavior modification (6, 7). A meta-analysis of self-help programs further demonstrated the effectiveness of the group-education approach (8). In a recent study of patients recruited from primary care practices, however, equivalent gains were made by patients attending group classes and controls who were instructed to read the class book (9).
The Stanford ASHC was recently adapted to meet the needs of the growing Spanish-speaking community that includes many immigrants with lower education levels. The Spanish adaptation of the ASHC was shown to improve exercise, disability, pain, and self efficacy after 4 months, and to maintain improvements after 1 year (10). The Orange County branch of the Arthritis Foundation, Southern California Chapter was licensed by Stanford University to present the Spanish ASHC as the Spanish Arthritis Empowerment Program (SAEP). The purpose of the present study is to evaluate whether the SAEP was successfully disseminated to another region of California and was effective among the low-income, indigent, and migrant laborers with arthritis in its Hispanic community.
- Top of page
- SUBJECTS AND METHODS
Participants in the SAEP classes were enrolled between October 1999 and May 2001. Of 216 people who enrolled in classes, 23 did not have arthritis, 9 were repeating, 16 were family members, and the remaining 168 were eligible for the study. Of those 168 eligible, 141 (84%) graduated and completed both pretest and posttest evaluations; 118 (83.7%) of these persons completed the 6-month followup evaluation. The 23 cases lost to followup were similar to the 118 who completed followup on all demographic characteristics. However, 22 of those lost to followup were from 1 site, suggesting an administrative or data collection problem.
Mean age of study participants (n = 141) was 50.7 years (SD 14.4) and 92.2% were women. Other characteristics of study participants are presented in Table 1. Approximately 80.9% had a noninflammatory type of arthritis, including osteoarthritis and low back pain, and 19.1% likely had inflammatory arthritis, including rheumatoid arthritis, lupus, and gout, as determined by patient self report of physician diagnosis of arthritis and mannequin joint count and pattern. However, only 6% had previously seen a rheumatologist regarding their arthritis.
Table 1. Baseline participant characteristics (n = 141)
|Nation of origin|| |
| Central America||10.0|
| South America||5.0|
|Education, years|| |
| 13 or more||6.4|
|Medical insurance|| |
|Employment status|| |
| Full/part time||32.6|
| Seeking work||5.7|
|Language spoken|| |
| Spanish only||60.0|
| Spanish with some English||40.0|
|Marital status|| |
|Current source of medical care|| |
| Private doctor/clinic||41.8|
| County hospital/clinic||25.5|
| Do not receive regular medical care||20.6|
|Occupation (current or former)|| |
| Service work (housekeeper, childcare)||18.4|
| None reported||26.2|
Significant improvement was seen from pretest to posttest in measures of disease activity using repeated-measures ANOVA (n = 141). The largest improvement was seen in the pain rating mean ± SD, which decreased from 5.6 ± 2.7 to 4.3 ± 2.5 (P = 0.0001). The patient self-reported tender and swollen joint counts decreased from a mean of 8.6 to 7.1 (P = 0.002) and 3.7 to 2.8 joints (P = 0.043), respectively. Function (mHAQ) did not significantly change from pretest (mean 0.52) to posttest (mean 0.51; P = 0.73). Patients reported significant improvement in mean pretest to posttest ratings (on a 0 to 10 scale) of arthritis impact on sleep impairment (4.1 to 3.1; P < 0.0001), daily activity impairment (4.8 to 3.4; P < 0.0001), and depression (4.6 to 3.8; P = 0.002).
Followup surveys were completed by 118 of the 141 program participants 6 months after enrollment. As shown in Table 2, the improvements at posttest were maintained at the 6-month followup. Although mHAQ was not significantly different from pretest to posttest, significant improvement was seen after 6 months. Overall, after 6 months, 0.9% rated their general health as excellent, 8.5% as very good, 31.4% good, 55.9% fair, and 3.3% poor. General health mean ± SD rating improved significantly from pretest (2.1 ± 0.7) to 6-month followup (2.5 ± 0.7; Z = 4.63, P < 0.0001); 33.9% improved, 60.2% remained the same, and only 5.9% reported worse health than at pretest.
Table 2. Repeated-measures analyses of variance comparing pretest to posttest and to 6-month followup (n = 118)*
|Measure||Pretest Mean ± SD||Posttest Mean ± SD||6-Month follow-up Mean ± SD||Repeated-measures planned contrasts|
|Disease activity measures|| || || || || || || |
| Pain (0–10 VAS)||5.99 ± 2.60||4.51 ± 2.51||3.44 ± 2.13||49.18||<0.0001||146.46||<0.0001|
| Tender joints (0–50)||9.22 ± 10.84||7.51 ± 9.70||6.72 ± 8.13||9.07||0.003||10.07||0.002|
| Swollen joints (0–48)||4.11 ± 8.51||3.13 ± 6.29||2.18 ± 5.32||3.36||0.069||9.99||0.002|
| mHAQ (0–3)||0.564 ± 0.688||0.539 ± 0.633||0.499 ± 0.608||1.63||0.205||5.21||0.024|
|Arthritis health measures (0–10 VAS, 10 is worst)|| || || || || || || |
| Sleep impairment||4.25 ± 3.76||3.09 ± 3.39||2.39 ± 2.71||27.25||<0.0001||53.67||<0.0001|
| Depression||4.75 ± 3.61||3.97 ± 3.29||2.96 ± 2.84||8.13||0.005||38.47||<0.0001|
| Daily activity impairment||5.05 ± 3.61||3.50 ± 3.16||2.49 ± 2.44||30.63||<0.0001||88.10||<0.0001|
|Self-management measures (0 is worst)|| || || || || || || |
| Self efficacy (0–10)||5.53 ± 3.57||7.84 ± 1.87||8.35 ± 1.57||57.10||<0.0001||79.82||<0.0001|
| Self-care behavior (0–8)||1.73 ± 1.40||4.07 ± 1.56||4.92 ± 1.50||249.6||<0.0001||397.5||<0.0001|
| Arthritis knowledge (0–7)||1.62 ± 1.59||4.42 ± 1.53||4.49 ± 1.31||198.8||<0.0001||209.8||<0.0001|
Stepdown linear regression analyses were done on 7 outcome variables to identify significant predictors of improvement (Tables 3 and 4). Dependent variables were change scores (pretest – 6-month), with positive scores indicating improvement. The pretest measure corresponding to each change score was included in the regression model for each outcome variable because it was expected that the greater the problem at pretest, the greater the improvement can be at followup. Self-management outcomes related to program content, e.g., self-efficacy score and self-care behavior and arthritis knowledge score at 6-month followup, were included as predictors of disease activity and arthritis health improvement. The mean ± SD number of physician visits in the 4 months prior to the pretest was 1.14 ± 1.9 and in the 4 months prior to the 6-month followup was 1.85 ± 1.5. The increase in physician visits was mainly due to 54 of 67 persons with no visits at pretest who then went to see a physician for their arthritis after the SAEP. Number of physician visits between posttest and 6-month followup was included in the regression model to see how physician visits contributed to improvement.
Table 3. Summary of regression models for each dependent variable (n = 118)*
|Dependent variable improvement score||R2||Adjusted R2||F||P||Constant + predictors included in each stepwise regression model|
|Pain improvement||0.509||0.501||59.67||<0.0001||Constant = −6.057 + pain (pretest) + efficacy (6 months)|
|No. tender joints improvement||0.524||0.512||41.88||<0.0001||Constant = −1.348 + no. tender joints (pretest) + inflammatory + no. physician visits|
|No. swollen joints improvement||0.806||0.802||238.5||<0.0001||Constant = −0.631 + no. swollen joints (pretest) + inflammatory|
|Impaired sleep improvement||0.593||0.582||54.40||<0.0001||Constant = −2.622 + sleep (pretest) + efficacy (6 months) + pain (6 months)|
|Depression improvement||0.612||0.601||58.83||<0.0001||Constant = −3.569 + depression (pretest) + efficacy (6 months) + pain (6 months)|
|Impaired daily activity improvement||0.684||0.676||80.97||<0.0001||Constant = −5.430 + daily activity (pretest) + efficacy (6 months) + pain (6 months)|
|mHAQ improvement||0.341||0.322||18.61||<0.0001||Constant = 0.116 + mHAQ (pretest) + no. physician visits (6 months) + pain (6 months)|
Table 4. Regression coefficients (B) and significance levels for predictors*
|Dependent variable||Pretest measure||Efficacy (6 months)||Self-care behavior/ knowledge (6 months)||No. physician visits (6 months)||Inflammatory arthritis||Pain rating (6 months)|
|Pain improvement||0.678||<0.0001||0.543||<0.0001||0.027||0.713||−0.048||0.489|| || || || |
|No. tender joints improvement||0.639||<0.0001||0.003||0.964||−0.001||0.992||−0.792||0.037||−7.65||0.001|| || |
|No. swollen joints improvement||0.886||<0.0001||0.046||0.299||−0.051||0.224||−0.058||0.185||−14.13||<0.0001|| || |
|Impaired sleep improvement||0.638||<0.0001||0.335||0.011||0.022||0.742||−0.006||0.924|| || ||−0.301||0.003|
|Depression improvement||0.719||<0.0001||0.421||0.004||0.071||0.280||0.018||0.773|| || ||−0.461||<0.0001|
|Impaired daily activity improvement||0.763||<0.0001||0.583||<0.0001||−0.104||0.077||−0.064||0.250|| || ||−0.217||0.020|
|mHAQ improvement||0.293||<0.0001||0.142||0.157||0.101||0.215||−0.036||0.034|| || ||−0.044||<0.0001|
The pretest measure was the best predictor of improvement score for each outcome (e.g., for pain improvement, pretest pain rating was the first predictor to enter the stepdown regression model with B = 0.678, P < 0.0001). After the pretest score, self efficacy was a significant predictor of improvement in 4 of 7 outcome variables (P < 0.0001) except the tender and swollen joint count and mHAQ scores. Having a type of inflammatory arthritis (by self report) was predictive of less improvement in number of tender and swollen joints (B = –7.65 and –14.13, respectively). Number of physician visits entered the regression model for improvement in tender joints and mHAQ. However, B was –0.792 for tender joints and –0.036 for mHAQ, suggesting higher improvements in tender joint count and mHAQ were related to fewer physician visits. Similarly, patients with lower improvement, e.g., worsening tender joint count or mHAQ score, had more physician visits. Pain rating entered the regression models for ratings of arthritis effect on sleep, mood, and daily activity, suggesting that pain is an important component of these outcomes. Attendance was very good; 52.5% attended all 6 classes, 39% attended 5, and the remaining 8.5% attended 4 of 6 classes. Attendance was not a significant predictor for any outcome measures.
- Top of page
- SUBJECTS AND METHODS
The SAEP as presented by the Arthritis Foundation, Southern California Chapter, in Orange County produced significant improvements in patient arthritis knowledge, self efficacy, general health, and arthritis symptoms from pretest to posttest, which were maintained over the 6-month followup period. Physical function (measured by the mHAQ) did not change from pretest to posttest, but showed improvement after 6 months. Similar findings were shown by Lorig et al (10) among Spanish-speaking patients in the Stanford area of northern California. Their sample was also predominantly from Mexico, with 46% having <7 years of education.
In contrast, Solomon et al (9) found no significant difference between intervention patients attending ASHC classes and controls who were given the ASHC book to read, and overall no significant changes from baseline to followup. These contrasting findings may be due to differences in the respective samples. Their English-speaking participants were recruited from a physician-hospital network, were older than our sample (mean age 68 years), 96% white, and well educated (29% of intervention and 49% control groups were college graduates). Many may have already been knowledgeable about pain control and coping skills at baseline. Moreover, all were under physician's care and 39% of intervention and 53% of controls were treated by a rheumatologist. Our sample had similar levels of pain and disability, but was on average 18 years younger and less well educated (55% had <7 years of education), with only 6% reporting care by a rheumatologist. Additional studies, including randomized controlled trials among various populations and settings, are needed to better target the ASHC to participants most likely to benefit.
Components of best practices in arthritis care should include effective patient education, such as the SAEP, which emphasizes empowerment and self-efficacy skills in coordination with appropriate physician care for arthritis. These components are difficult to achieve in lower-educated, indigent populations with less access both to health information and health care. The SAEP empowered participants by improving self-management measures (self efficacy, self-care behavior, and arthritis knowledge), which contributed to the significant improvements in measured outcomes at 6-month followup. Empowerment may include self-management skills as well as learning how to gain access to health care, finding appropriate physician care, and communicating more effectively with their physicians. As evidenced by the linear regression models to predict change in tender joint count and mHAQ, patients with worsening scores tended to have an increased number of physician visits, suggesting that the SAEP may have empowered them to seek physician care as appropriately indicated.
In summary, the Spanish Arthritis Empowerment Program was successfully disseminated and implemented in a lower-educated Spanish-speaking population in Southern California. Participants improved their self-management skills, knowledge of arthritis, appropriate seeking of health care for arthritis, and implemented changes in self-care behavior related to arthritis.