Patient-reported antidepressant discontinuation after adverse events and associated factors: A retrospective...

To investigate the patient-reported antidepressant discontinuation rate after adverse events (AEs) and its associated factors.

This retrospective recall-based cross-sectional survey was conducted at Beijing Anding Hospital from April 7, 2025, to January 21, 2026. A total of 500 adult patients (≥18 years) who had used antidepressants and experienced AEs within the past year were enrolled and completed structured face-to-face questionnaires. The primary outcomes were the AE-related discontinuation rate (AE-DR, event level) and the discontinuation proportion (DP, patient level). Chi-square tests compared AE-DR across antidepressant classes and organ systems. Cluster bootstrap-corrected multivariate generalized linear mixed models (GLMMs) identified independent correlates of discontinuation.

Among 500 patients, 905 AEs were reported. The AE-DR was 38.6% (349/905), and the DP was 57.0% (285/500). Significant differences in AE-DR were observed across antidepressant classes (P = 0.038) and organ systems (P < 0.001). NaSSAs showed the highest AE-DR (77.8%, 14/18), followed by melatonin receptor agonists (64.3%, 9/14) and SARIs (54.8%, 40/73). The reproductive system had the highest AE-DR (96.8%, 60/62), followed by the hematological (88.9%, 8/9) and metabolic systems (73.2%, 52/71). The gastrointestinal tract accounted for the largest number of AEs (281) but had the lowest discontinuation rate (15.3%, 43/281). Cluster bootstrap-corrected GLMM analysis revealed that greater AE severity, reproductive system involvement, SARI antidepressants, psychosocial factors (insufficient family support, high occupational stress, higher monthly household income), and clinical characteristics (greater number of comorbid mental disorders, more episodes) were all independent correlates of AE-related discontinuation (all P < 0.05). Family support was the core protective factor.

This cross-sectional study based on structured patient self-reported data identified the post-AE discontinuation rate and its independent correlates. AE severity, specific organ systems (particularly the reproductive system), and patient psychosocial and clinical characteristics were independently associated with discontinuation. When AEs occur, it is necessary for clinicians to carefully monitor and manage patients with these characteristics. Notably, these findings primarily reflect the post-AE discontinuation rate and its correlates rather than the overall discontinuation rate in the general clinical population. Due to self-report bias, a single-center setting, and limited sample sizes for certain subgroups, these findings require further validation in larger, independent cohorts.