Stigma Associated With Mental Illness: A Cross-strait Comparison Between Shanghai And Taiwan

[Objective]We aimed to apply Shanghai's "Assessment Scale for Stigma towards Mental Illnesses" and"Assessment Scale for Discrimination against Patients with Mental Illnesses" in the Taiwan region. The results of these scales were evaluated for reliability and applicability. We aim to better understand the stigma level of different Taiwanese population groups; as well as to examine for any possible difference in stigma level between respective Shanghai and Taiwan population groups. We also analyzed characteristics that may contribute to mental illness stigma.[Methods]The study employed the "Assessment Scale for Stigma towards Mental Illnesses" and"Assessment Scale for Discrimination against Patients with Mental Illnesses" in a sample of 169 in-patients and out-patients with mental illnesses, as well as 497 other participants (including the psychiatric staff, families and friends of patients with mental illness, and general public) in Taiwan. Thirty psychiatric staff then re-took the test after a week to evaluate the test-retest reliability. Participants were approached by investigators and they completed the written questionnaire autonomously. The "Assessment Scale for Discrimination against Patients with Mental Illness" was also posted on the World Wide Web to allow the general population to complete the online version of the questionnaire. The content of the questionnaire consisted of demographic information, factors related to mental illnesses, and stigma level assessment. Factors related to mental illnesses included enquiries such as previous participation in psychiatric/psychological seminars, and the extent of contact with people with mental illnesses. In stigma level assessment, participants select answers according to each entry; answer options ranged from "regularly" to "never", and from "completely agree" to "completely disagree". Answer options were rated using the Likert scale scoring system, and the scale score is calculated and analyzed accordingly.Factor analysis, internal consistency test and split-half reliability test were then performed on the results of the above two scales. Test-retest reliability was tested a week after for the "Assessment Scale for Discrimination against Patients with Mental Illness". F test was performed to explore between-groups difference in different Taiwanese participant groups in terms of level of discrimination against mental illnesses. The results of the Shanghai region have preciously been completed and published. Results from the Taiwan and Shanghai regions were compared with t test to explore differences in discrimination level between respective groups in the two localities. Lastly, t test and F test were performed on demographic information and factors related to mental illnesses to explore the influences of these variables on the level of discrimination.[Results]1. Three factors were extracted from the "Assessment Scale for Stigma towards Mental Illnesses", namely exclusion of others, disease shamefulness and social functions; they comprised 37%, 9% and 6%, respectively, of the total variance. Loading values were between 0.24-0.82. The Cronbach's a coefficient and split-half reliability coefficients of the total scale were 0.93 and 0.85, respectively. The three factors'Cronbach's a coefficient were in the range of 0.48-0.93; the split-half reliability coefficient of the exclusion of others factor was 0.89.2. Three factors were extracted from the "Assessment Scale for Discrimination against Patients with Mental Illnesses", namely belittlement, communication and social isolation; they comprised 32%, 7% and 6%, respectively, of the total variance. Loading values were between 0.40 and 0.77. The Cronbach's a coefficient and split-half reliability coefficients of the total scale were 0.91 and 0.88, respectively. For the belittlement factor, Cronbach's a factors and split-half reliability coefficients were 0.91 and 0.89, respectively. The remaining two factors had low a coefficients; and due to the small number of question items, split-half reliability test was not conducted. Test-retest reliability demonstrated that the correlation coefficient between the two tests a week apart was 0.86. For the belittlement factor, correlation coefficient was 0.83. The remaining two factors had low correlation coefficients.3. There were statistically significant difference in the extent of discrimination among different Taiwanese population groups (F=6.892, p<0.005). In particular, the psychiatric staff (discrimination score 42.58±14.891) showed significantly lower extent than the other groups (discrimination score of families and friends of ward and clinic patients =49.18±16.358; families and friends of the general public=48.65±14.380; general public=51.60±16.327). When comparing the extent of discrimination between different Shanghai and Taiwan population groups, it was shown that Taiwanese psychiatric staff (discrimination score 42.58±14.891) had substantially lower discrimination level than their Shanghai's counterparts (discrimination score=48.15±16.587), with F=1.558,p<0.005. Further, Taiwanese general public (discrimination score=51.60±16.327)also had lower discrimination level than Shanghai's (discrimination score=58.71±15.332), with F=0.107, p<0.005. However, families and friends of patients with mental illnesses showed no apparent difference in the two localities (Discrimination score for Taiwan and Shanghai were 48.84±15.038 and 47.91±17.266, respectively).4. In analysis of characteristic factors, statistically significant differences were obtained when comparisons were made according to whether samples "have previously attended psychiatric/psychological seminars". Those sample who have previously attended psychiatric/psychological seminars have a discrimination level of 47.75±15.760; those who have not have a discrimination level of 52.50±15.692, with F=10.330,p<0.005. When comparing"the extent of contact with mental illness", the difference in discrimination score was also statistically significance. Discrimination scores was 48.95±15.432 for samples with contact level <7,38.06±16.188 for samples with contact level=7, and 39.88±15.918 for samples with contact level>7 (F=5.937, p<0.005). The difference in discrimination score was also statistically significance when comparing education levels. No statistically significant differences were obtained in other characteristic factors.[Conclusion]1. The two scales developed in Shanghai showed good reliable when applied in the Taiwan region. These two scales are suitable for the use in the Taiwan region as an assessment tool for mental illness stigma and discrimination level. This provided grounds for further investigation to be done on this field.2. Discrimination levels of various Taiwanese population groups were statistically different. This may imply that contact with and understanding of mental illness can influence discrimination level; but the exact causes of this difference needs to be ascertained with further studies.3. There were significant differences in discrimination level between respective Shanghai and Taiwan population groups. This difference may be accounted by the difference in health care resource distribution and in health-care related policies. Differences in community education strategies may also contributed to these differences between the two localities.4. Certain characteristics related to mental illness can have substantial impact on the extent of discrimination. Public education and instruction schemes could be useful means of increasing understanding and awareness of mental illnesses, and possibly improve issues associated with discrimination.