Discrepancies between pre-specified and reported primary outcomes: A cross-sectional analysis of randomized controlled trials in gastroenterology and hepatology journals

Abstract

Background: Previous research has raised concerns regarding inconsistencies between reported and pre-specified outcomes in randomized controlled trials (RCTs) across various biomedical disciplines. However, studies examining whether similar discrepancies exist in RCTs focusing on gastrointestinal and liver diseases are limited. This study aimed to assess the extent of discrepancies between registered and published primary outcomes in RCTs featured in journals specializing in gastroenterology and hepatology.

Methods: We retrospectively retrieved RCTs published between January 1, 2017 and December 31, 2021 in the top three journals from each quartile ranking of the 2020 Journal Citation Reports within the "Gastroenterology and Hepatology" subcategory. We extracted data on trial characteristics, registration details, and pre-specified versus published primary outcomes. Pre-specified primary outcomes were retrieved from the World Health Organization's International Clinical Trials Registry Platform. Only trials reporting specific primary outcomes were included in analyzing primary outcome discrepancies. We also assessed whether there was a potential reporting bias that deemed to favor statistically significant outcomes. Statistical analyses included chi-square tests, Fisher's exact tests, univariate analyses, and logistic regression.

Results: Of 362 articles identified, 312 (86.2%) were registered, and 79.8% of the registrations (249 out of 312) were prospective. Among the 285 trials reporting primary outcomes, 76 (26.7%) exhibited at least one discrepancy between registered and published primary outcomes. The most common discrepancies included different assessment times for the primary outcome (n = 32, 42.1%), omitting the registered primary outcome in publications (n = 21, 27.6%), and reporting the registered secondary outcomes as primary outcomes (n = 13, 17.1%). Univariate analyses revealed that primary outcome discrepancies were lower in the publication year 2020 compared to year 2021 (OR = 0.267, 95% CI: 0.101, 0.706, p = 0.008). Among the 76 studies with primary outcome discrepancies, 20 (26.3%) studies were retrospectively registered, and 32 (57.1%) of the prospectively registered trials with primary outcome discrepancies showed statistically significant results. However, no significant differences were found between journal quartiles regarding primary outcome consistency and potential reporting bias (p = 0.14 and p = 0.28, respectively).

Conclusions: This study highlights the disparities between registered and published primary outcomes in RCTs within gastroenterology and hepatology journals. Attention to factors such as the timing of primary outcome assessments in published trials and the consistency between registered and published primary outcomes is crucial. Enhanced scrutiny from journal editors and peer reviewers during the review process is necessary to ensure the reliability of gastrointestinal and hepatic trials.

Fig 1. Flow diagram of identification of journals and articles.