TY - JOUR
T1 - Meaningful and accurate disclosure of conflict of interest at the ASTRo national meeting
T2 - A need for reassessment of current policies
AU - Ahmed, Awad A.
AU - Yoo, Stella K.
AU - Mehta, Shahil
AU - Holliday, Emma B.
AU - Deville, Curtiland
AU - Vapiwala, Neha
AU - Wilson, Lynn D.
AU - Jagsi, Reshma
AU - Prasad, Vinay
AU - Thomas, Charles R.
N1 - Funding Information:
Conflicts of interest (COIs) by researchers and clinicians in medicine have been the subject of numerous reports.1-4 Although a COI can be any situation or relationship in which a physician may be biased because of personal gain, financial relationships between health care professionals and industry sponsors are particularly scrutinized because of their potentially lucrative nature. These relationships can render research susceptible to errors in methodology.5 They can also bias outcomes by leading to the withholding of negative results5 and producing outcomes that favor the sponsor. This can, in turn, adversely affect medical decision making and standards of care.5 Academic societies and journals have policies that promote full disclosure of any potential COIs,6 and although large efforts have been made by governmental organizations, universities, and professional societies to better address COIs, it is important to acknowledge that disclosure is not a panacea for COIs in medicine.7 Generally, COIs have been addressed by requiring stricter disclosures policies; however, policies requiring the removal of the COI and/or selective exclusion of medical professionals from research related to COIs have also been implemented.8 For example, institutions that received COI disclosures from researchers were annually required to evaluate and report to the National Institutes of Health whether COIs could affect the researcher’s work and determine the extent to which investigators needed to disclose COIs in publications or limit their involvement in projects.9 ASCO also identifies financial relationships that are problematic and cannot merely be managed by disclosure. Specifically, explicit restrictions are placed on committee members involved in guideline recommendations; these include forbidding the majority of panel members, including the panel chair, from holding relationships with relevant companies, as well as not accepting industry funding toward clinical practice guidelines.6 Similarly, the Association of American Medical Colleges (AAMC) prohibits financial relationships in research when potential conflicts may pose a significant risk to human participants. Previous studies have examined the accuracy of disclosure of COI statements at medical conferences10 and whether disclosure protocols establish useful disclosure.11 The medical field is moving toward preventing biases from financial COIs by educating medical students and encouraging development of guidelines in professional medical associations.12,13 In an effort to improve disclosures, the Centers for Medicare and Medicaid Services has instituted an online public database for reporting of payments made to US physicians as part of the Physician Payments Sunshine Act.14 Readability of COI slides at the ASCO national meeting was examined using 4 words per second (wps) as the threshold for readability (ie, the speed representing the cutoff above which most readers are unlikely to comprehend what they are reading).11 We sought to examine the nature and accuracy of COI disclosure at the American Society for Radiation Oncology (ASTRO) national meeting over the last 3 years to determine whether disclosures are displayed at a speed adequate for comprehension for most readers and the accuracy of disclosures, as determined by cross-referencing the Centers for Medicare and Medicaid Services Open Payments database.
Publisher Copyright:
Copyright © 2018 by American Society of Clinical Oncology.
PY - 2018/2
Y1 - 2018/2
N2 - Purpose Conflict of interest (COI) disclosure is essential to research integrity. The average reading comprehension in English is 3.8 words per second (wps). This study examines presenters at the American Society for Radiation Oncology (ASTRO) national meeting over a recent 3-year period to determine whether disclosure is presented accurately and in a manner that allows the audience to comprehend the content. Methods We examined videos of presentations as well as slides from 2014 to 2016 from the ASTRO virtual meeting, noting whether a COI slide was presented, the duration the slide was visible, and the number of disclosures. Disclosures were cross-referenced for discrepancies with the publicly reported Centers for Medicare and Medicaid Services Open Payments database. Using a cutoff of 4 wps, we noted how many presentations were presented at speeds of # 4 wps and . 4 wps. Results The final data set consisted of 401 presentations delivered by 364 presenters. Using a threshold of 4 wps, 34.0% of presenters had COI slides shown too fast for the average audience to comprehend. Moreover, 16.3% of US physicians incorrectly underreported industry funding received. Of these presentations with discrepancies, 32.6% did not have a COI slide, 39.5% failed to disclose any COI, 27.9% partially disclosed COIs, and 11.6% contained multiple discrepancies. The number of wps were correlated with having a discrepancy on multivariable regression (P = .046; odds ratio, 1.08; 95% CI, 1.01 to 1.19). Conclusion A substantial minority of presentations at ASTRO lack meaningful disclosure, and a surprising number incorrectly reported COIs. Additional guidance may be needed to promote more meaningful and accurate disclosure of COIs at major national meetings in oncology.
AB - Purpose Conflict of interest (COI) disclosure is essential to research integrity. The average reading comprehension in English is 3.8 words per second (wps). This study examines presenters at the American Society for Radiation Oncology (ASTRO) national meeting over a recent 3-year period to determine whether disclosure is presented accurately and in a manner that allows the audience to comprehend the content. Methods We examined videos of presentations as well as slides from 2014 to 2016 from the ASTRO virtual meeting, noting whether a COI slide was presented, the duration the slide was visible, and the number of disclosures. Disclosures were cross-referenced for discrepancies with the publicly reported Centers for Medicare and Medicaid Services Open Payments database. Using a cutoff of 4 wps, we noted how many presentations were presented at speeds of # 4 wps and . 4 wps. Results The final data set consisted of 401 presentations delivered by 364 presenters. Using a threshold of 4 wps, 34.0% of presenters had COI slides shown too fast for the average audience to comprehend. Moreover, 16.3% of US physicians incorrectly underreported industry funding received. Of these presentations with discrepancies, 32.6% did not have a COI slide, 39.5% failed to disclose any COI, 27.9% partially disclosed COIs, and 11.6% contained multiple discrepancies. The number of wps were correlated with having a discrepancy on multivariable regression (P = .046; odds ratio, 1.08; 95% CI, 1.01 to 1.19). Conclusion A substantial minority of presentations at ASTRO lack meaningful disclosure, and a surprising number incorrectly reported COIs. Additional guidance may be needed to promote more meaningful and accurate disclosure of COIs at major national meetings in oncology.
UR - http://www.scopus.com/inward/record.url?scp=85068807223&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85068807223&partnerID=8YFLogxK
U2 - 10.1200/JOP.18.00121
DO - 10.1200/JOP.18.00121
M3 - Article
AN - SCOPUS:85068807223
VL - 14
SP - e692-e698
JO - Journal of Oncology Practice
JF - Journal of Oncology Practice
SN - 1554-7477
IS - 11
ER -