Social stigma is inevitable for mentally ill patients, but how patients treat themselves is a priority for rehabilitation and an important buffer mechanism. This study thus aimed to measure the effectiveness of rehabilitation models for improving self-stigma. This quasi-experimental research design applied purposeful sampling. The participants (n = 250) were persons with mental illness who received rehabilitation treatment in central Taiwan. They were divided into community- (n = 170) and institution-based (n = 80) rehabilitation groups. The Internalized Stigma of Mental Illness Scale was evaluated at the time of recruitment, and a follow-up was conducted after 1 and 3 months. A generalized estimation equation was used in data analysis to measure whether self-stigma improved with the rehabilitation model and time, and to test the effect of different rehabilitation models on participants' self-stigma improvement. The study found that the self-stigma of patients receiving CBR improved more than that of those receiving IBR when behavioral problems, education, OT level, sex, and first-time self-stigma were controlled. Returning to the community is the goal of rehabilitation for patients with mental illness, but IBR still dominates the rehabilitation model in Taiwan. Thus, it is necessary to continue promoting CBR plans for future mental health policies.
Keywords: community; generalized estimation equation (GEE); institution; mentally ill; rehabilitation; self-stigma; stigma.