Medication Adherence in Severe Mental Illness: a Promotion Program

Bipolar Disorder (BD), Schizophrenia Disorder, Severe Mental Disorders, Treatment Adherence and Compliance

treatment adherence, schizophrenia, bipolar disorder, electronic health records, medication

People living with serious mental illnesses such as schizophrenia and bipolar disorder often need long-term medication to stay well. However, many patients have difficulty taking medication regularly, which can increase the risk of relapse, hospitalization, and poorer quality of life. Traditionally, treatment adherence has been measured using self-report questionnaires, which may be influenced by memory or social desirability bias. With the recent expansion of electronic prescription systems in Spain, it is now possible to objectively verify whether patients collect their medications from the pharmacy. This provides a new opportunity to better understand and support treatment adherence. The ADHERA study will evaluate how well digital self-report questionnaires reflect real medication use compared with electronic dispensing records. We will also explore patient characteristics that may be associated with difficulties in medication adherence. Finally, we will test a new online psychoeducational program-including sessions led by mental health professionals and supported by peer-experience contributors-to determine whether it can help improve adherence. Participants with schizophrenia or bipolar disorder who are registered in the hospital's digital patient portal and have active antipsychotic prescriptions will be invited to complete brief adherence questionnaires online. Individuals with signs of reduced adherence will then be invited to take part in a telehealth intervention consisting of ten group sessions, where they will receive information, support, and practical strategies to maintain their treatment plan. Medication adherence will be reassessed after six months. If successful, this study may help improve how treatment adherence is measured in clinical practice, guide targeted interventions for individuals at higher risk of non-adherence, and provide evidence for scalable telehealth programs that can be easily implemented in other regions and medical conditions

This study aims to evaluate the validity of digital self-reported treatment adherence questionnaires in comparison with objective pharmacy dispensing records for antipsychotic medications, identify predictors of non-adherence, and assess the effectiveness of a telehealth intervention designed to improve adherence among adults with schizophrenia or bipolar disorder. Study Design The protocol incorporates three linked components: Validation study: Sensitivity, specificity, predictive values, and ROC analysis will be used to determine the accuracy of digital adherence self-report scales compared with electronic dispensing data from the regional Single Prescription Module (MUP), considered the reference standard. Observational study: Sociodemographic and clinical variables obtained through the Patient Portal will be analyzed to identify correlates of adherence. Logistic regression and correlation models will be applied to examine associations between treatment exposure (continuous and categorical measures) and relevant predictors. Pre-post intervention: Participants identified as having poor adherence will be offered a structured psychoeducational program delivered via videoconference. The program includes ten group sessions led by clinical psychologists and two peer-contributor sessions, focusing on knowledge reinforcement, self-management skills, and barriers to adherence. Medication exposure levels and self-report scores will be reassessed six months post-intervention. Setting and Population Participants will be recruited through four public hospitals in the Community of Madrid. Eligible individuals must be adults diagnosed with schizophrenia or bipolar disorder, registered in the digital patient portal, and have active antipsychotic prescriptions. Patients unable to provide informed consent or without portal access will be excluded. Consent will be obtained electronically. The estimated eligible population is approximately 4,900 individuals. Expected questionnaire completion is \ 30%, yielding \ 1,640 participants for the validation component. Based on expected non-adherence rates (40-50%), approximately 650 participants are anticipated for the intervention phase. Data Sources and Security Adherence outcomes will include digital self-report measures (Morisky-Green and Haynes-Sackett) and percentage of antipsychotic exposure obtained from the MUP over the prior six-month period. Sociodemographic and clinical characteristics (age, sex, urbanicity, comorbidities) will be extracted from the Patient Portal. Personal identifiers will be pseudonymized and stored separately from study records. Access to data will be restricted to authorized study personnel. Digital infrastructure includes encrypted communication, firewalls, intrusion-detection systems, and periodic audit mechanisms. Data processing follows GDPR and Spanish data protection laws. Statistical Plan Analyses will employ SPSS and R. For the validation phase, diagnostic accuracy metrics and ROC curves will be calculated. For risk-factor analysis, logistic regression (using \<75% exposure as a risk threshold) and correlation analyses will be performed. For intervention evaluation, pre-post comparisons of pharmacological exposure and questionnaire scores will be conducted using appropriate paired statistical tests. Risks and Burden The intervention consists of group psychoeducational telehealth sessions, with no known risks. Participants will not receive financial incentives. Data confidentiality and voluntary participation are ensured. Study Relevance To our knowledge, this is the first Spanish study validating digital adherence questionnaires against real-world electronic dispensing data in patients with severe mental illness, and one of the first to integrate digital phenotyping with a peer-supported telehealth adherence program. The findings may support scalable, patient-centered models to improve adherence and long-term outcomes.

No related papers found yet.