Factors associated with medication experience in patients with schizophrenia: a mixed-methods study across three post-discharge time points

Abstract

Background Schizophrenia requires sustained antipsychotic therapy, yet medication adherence remains suboptimal in over 50% of patients due to suboptimal medication experiences, often leading to self-directed dose reduction or discontinuation. This non-adherence contributes to disease relapse, functional impairment, and substantial familial economic burdens, while recurrent hospitalizations strain healthcare systems. We investigated correlates of medication experiences in patients with schizophrenia using a mixed-methods design with repeated assessments across three post-discharge time points. Methods Our cohort comprised 425 clinically stable patients with schizophrenia from two tertiary psychiatric hospitals in Heilongjiang Province. Using a mixed-methods design with repeated assessments at three post-discharge time points, we collected concurrent quantitative and qualitative data at three timepoints: pre-discharge (T1), 1-month (T2), and 3-months post-discharge (T3). Quantitative measures included standardized assessments of medication experience (SWN), burden (LMQ), beliefs (BMQ), symptomatology (PANSS), social support (SSRS), and family functioning (APGAR). Machine learning models (logistic regression, random forest, support vector machine) identified key factors associated with medication experience. Qualitative data from 15 patients with suboptimal experiences were analyzed thematically, with mixed-methods integration. Results Random Forest models consistently demonstrated superior predictive performance, identifying the top predictors for each time-point: at T1 - medication beliefs, medication burden and psychotic symptoms (in descending importance); at T2 - medication burden, beliefs, family functioning, and symptoms; at T3 - medication beliefs, social support, burden, symptoms, and family functioning. Qualitative analysis revealed four core themes at T1 and three at T2/T3, showing complementary quantitative-qualitative relationships in benefit-risk perceptions and treatment valuation. Conclusions These findings suggest psychiatric nurses may consider interventions that address modifiable factors associated with medication experience: mitigating medication burdens while enhancing perceived treatment benefits during hospitalization (T1); strengthening medication beliefs and family support early post-discharge (T2); and bolstering social support networks during community reintegration (T3). These findings help identify potential intervention targets for future practice to optimize medication experience, and may have implications for medication-related clinical outcomes in patients with schizophrenia.