Protocol for a randomised controlled trial on smartphone- and web-based app for precision dosing and monitoring of lithium («PharmOracle») for improving adherence to therapy in patients with bipolar affective disorder or schizoaffective disorder

Abstract

Lithium remains a first-line treatment for bipolar disorder but its use is declining owing to the need for intensive monitoring and concerns about tolerability. Digital decision-support tools that combine model-informed precision dosing, Bayesian therapeutic-drug-monitoring, and adherence support may facilitate safer, more effective lithium therapy and improve long-term adherence. We aimed to evaluate whether adding a mobile application with precision-dosing, monitoring reminders and medication-support functions («PharmOracle») to standard care reduces a composite measure of non-adherence and adverse outcomes over 12 months, compared with standard care alone. We planning an open-label, multicenter, randomized (1:1) superiority trial enrolling adults (≥18 years) with ICD-10 bipolar I/II or schizoaffective disorder who are initiating lithium carbonate. Planned sample size is 210 participants (105 per arm). The primary endpoint is time to first occurrence of a composite outcome: lithium discontinuation for any reason, hospitalization for lithium toxicity (serum lithium ≥1.2 mmol/L with symptoms), or unplanned psychiatric hospitalization/emergency visit for mood relapse assessed continuously and verified at 3, 6 and 12 months. Key secondary endpoints include changes in the Integral Diagnostic Index for Bipolar Disorder composite index for assessing symptoms of depression (HDRS), mania (YMRS), psychosocial functioning (PSP), and quality of life (WHOQOL), self-reported medication adherence (10-item Medication Adherence Report), adherence to recommended laboratory/instrumental monitoring, proportion of patients achieving therapeutic serum lithium (0.6-1.0 mmol/L), and safety outcomes. The intervention comprises a mobile/web platform performing model-informed individualized dose prediction, Bayesian dose refinement from measured troughs, automated monitoring reminders, adverse-event self-reporting, and clinician dashboards. Analyses will follow intention-to-treat principles using survival methods (Kaplan-Meier, log-rank, Cox regression) for the primary outcome and mixed models or appropriate nonparametric tests for longitudinal secondary outcomes. We anticipate that using of mobile application with precision-dosing, monitoring reminders and medication-support functions will reduce the incidence of the composite outcome by improving adherence and accelerating attainment of therapeutic lithium concentrations, while maintaining an acceptable safety profile.

More by Aleksandr Sinitca

• Seborrheic dermatitis: Identification of key disease factors and treatment with a topical solution of betamethasone 0.05% + salicylic acid 2%Meditsinskiy sovet = Medical Council · 2026
• Reconstruction of long-term complex system dynamics from short-term fragments of observational records using high-order Kalman filteringChaos An Interdisciplinary Journal of Nonlinear Science · 2026
• Microscopy Image Dataset of pulmonary vessels for Quantitative assessment of fibrosisScientific Data · 2026
• Microscopy Image Dataset of pulmonary vessels for Deep Learning-Based Quantitative assessment of fibrosisFigshare · 2026