Abstract 164: Discontinuation and Non‐Publication of Stroke Clinical Studies: A Cross‐Sectional Analysis

Background Stroke is the second leading cause of both disability and death worldwide, with the highest burden of the disease shared by low‐ and middle‐income countries. In 2022, there are over 101 million people currently living who have experienced stroke globally. Clinical studies in stroke are vulnerable to discontinuation and non‐publication, representing significant sources of research waste in clinical medicine. This study aims to assess stroke clinical studies characteristics and summarize logistic, financial, and practical reasons behind early discontinuation and non‐publications to prevent this in the future. Methods ClinicalTrials.gov was searched for all studies registered between 1 January 2000 and 1 May 2022 and included patients with stroke. Publications from these studies were identified by extensive online searching using the NCT identifier number and other related keywords. Multiple logistic regression analysis was performed to identify characteristics associated with trial discontinuation and non‐publication. Results A total of 3720 eligible registered studies were included; 2724 (73.3%) were clinical trials, and 994 (26.7%) were observational studies. Of these, 380 (10.2%) were discontinued. Of the completed studies, there were a total of 1786 (53.5%) non‐published studies. The unadjusted logistic regression analysis revealed that the industrial (OR = 1.37, 95% CI [1.05‐1.79]) and mixed funding (OR = 1.42, 95% CI [1.04‐1.92]) were significant predictors of study discontinuation compared to non‐industrial funding. Furthermore, the small sample size (less than 100) (OR = 2.79, 95% CI [2.13‐3.66]) was also significant. Concerning the non‐publication, a small sample size (less than 100) (OR = 1.68, 95% CI [1.46‐1.93]) was also a significant predictor, while mixed funding (OR = 0,56, 95% CI [0.45‐0.70]) compared to non‐industrial funding has significantly decreased the likelihood of non‐publication. Behavioral interventions were significantly less likely to be discontinued than drug interventions (OR = 0,40, 95% CI [0.27‐0.60]). The Diagnostic interventions were significantly the most likely to be non‐published compared to drug interventions (OR = 3.39, 95% CI [1.85‐6.21]). Conclusion There is evidence of non‐dissemination bias in clinical studies of stroke. These biases distort the therapeutic information available to inform clinical practice and raise ethical concerns regarding exposing volunteering participants to potential risks without furthering practice.