Clinical validation of a composite medical device combining blood biomarkers, a decision algorithm, and the Mood Disorder Questionnaire to support the differential diagnosis of bipolar disorder in primary-care patients with a major depressive episode

Abstract Background Bipolar disorder is common, highly disabling and frequently misdiagnosed as unipolar depression in primary care, leading to inappropriate antidepressant treatment and substantial diagnostic delays. Early and accurate identification of bipolar disorder in adults presenting with a major depressive episode could improve prognosis, reduce suicide risk and lower healthcare and societal costs. Several questionnaires and some biomarkers a have been proposed to bridge the gap to a better diagnosis accuracy. Methods This is a multicenter, prospective study conducted in France. General practitioners (GPs) and/or primary care centers prescreen adults aged 18–65 years with a major depressive episode, as determined by their clinical assessment, who are eligible for antidepressant treatment. After referral, participants attend a specialized outpatient psychiatric center where they undergo a structured diagnostic interview (MINI), clinician‑ and self‑rated scales, and fasting blood sampling. A composite in vitro diagnostic algorithm integrating serum immune‑inflammatory biomarkers (including IL‑6, IL‑10, IL‑15, IL‑27 and CXCL10), clinical variables (age, psychotropic treatment) and the Mood Disorder Questionnaire (MDQ) is applied. Its diagnostic performance for bipolar disorder (sensitivity, specificity, predictive values, area under the Receiver Operating Characteristic (ROC) curve) is assessed against an expert Diagnostic and Statistical Manual-5 (DSM‑5) based diagnosis at baseline and at 7–9‑week follow‑up. The planned sample size is 623 participants, providing adequate precision for sensitivity and specificity estimates under an expected bipolar disorder prevalence of 10% in this population. Secondary objectives include characterization of false positives and false negatives, assessment of prognostic performance for hypomanic switch and depressive improvement at 8 weeks, evaluation of analytical variability and acceptability, comparison with GP diagnostic judgement. Acceptability study and medico‑economic modelling using the French national health database (SNDS) are planned as parallel studies. Discussion This trial will test whether a composite diagnostic medical device combining blood biomarkers and the MDQ improves bipolar disorder screening compared with the MDQ alone in a real‑world primary‑care pathway. If validated, the device could support earlier and more precise diagnosis, inform treatment decisions and provide data to model its impact on care pathways and costs. Trial registration Clinicaltrials.gov. Number: NCT07031817. Registered May 30, 2025