Search

Financial relationships with industry may bias educational content delivered by physicians. SAGES strives to mitigate potential bias, relying on physician self-reporting. Retrospective review of relationships is possible using the Open Payments Database (OPD), a public record of industry-reported payments to US physicians. We aimed to evaluate the effectiveness of the SAGES disclosure process by comparing faculty disclosures to SAGES, faculty disclosures within presentations, and OPD records among speakers at the 2018-2020 SAGES meetings.We reviewed all presentations from the SAGES 2018-2020 Annual Meetings. For each invited presentation, all slide-disclosed relationships were recorded. For US physicians, we queried the OPD and recorded relationships ≥ $500 USD in the calendar year prior to presentation. We compared the slide-disclosed relationships with OPD-reported relationships and with those provided to SAGES during the faculty disclosure process. We surveyed a sample of the 2020 annual meeting speakers to analyze potential reasons for discordance.From 2018 to 2020, there were 1,355 invited presentations, of which 1,234 (91%) were available for review. Disclosure slides were present in 1,098 (89%), increasing from 86% in 2018 to 93% in 2020. The proportion of speakers with OPD-reported relationships ≥ $500 increased from 54% in 2018 to 66% in 2020. The total value of OPD relationships decreased from $5.9 million (2018) to $3.3 million (2020) with a concomitant decrease in the proportion with high discordance from 9% in 2018 to 5% in 2020. Among the 2020 speakers with high discordance, the most common explanations for discordance were being unaware of payment or payment outside the 12-month timeframe (55%).Discordance between financial disclosures reported to SAGES and OPD highlight the need for improvements in the faculty disclosure process. SAGES will continue to streamline this process by incorporating faculty review of their OPD disclosures to ensure all educational programs remain free of commercial bias.