Hani A Alhamoud1, Ramzi Dudum1, Heather A Young2, Brian G Choi3. 1. The George Washington University School of Medicine and Health Sciences, Washington, DC. 2. The George Washington University Milken Institute School of Public Health, Washington, DC. 3. The George Washington University School of Medicine and Health Sciences, Washington, DC. Electronic address: bchoi@gwu.edu.
Abstract
BACKGROUND: Industry manufacturers are required by the Sunshine Act to disclose payments to physicians. These data recently became publicly available, but some manufacturers prereleased their data since 2009. We tested the hypotheses that there would be discrepancies between manufacturers' and physicians' disclosures. METHODS: The financial disclosures by authors of all 39 American College of Cardiology and American Heart Association guidelines between 2009 and 2012 were matched to the public disclosures of 15 pharmaceutical companies during that same period. Duplicate authors across guidelines were assessed independently. Per the guidelines, payments <$10,000 are modest and ≥$10,000 are significant. Agreement was determined using a κ statistic; Fisher's exact and Mann-Whitney tests were used to detect statistical significance. RESULTS: The overall agreement between author and company disclosure was poor (κ = 0.238). There was a significant difference in error rates of disclosure among companies and authors (P = .019). Of disclosures by authors, companies failed to match them with an error rate of 71.6%. Of disclosures by companies, authors failed to match them with an error rate of 54.7%. CONCLUSIONS: Our analysis shows a concerning level of disagreement between guideline authors' and pharmaceutical companies' disclosures. Without ability for physicians to challenge reports, it is unclear whether these discrepancies reflect undisclosed relationships with industry or errors in reporting, and caution should be advised in interpretation of data from the Sunshine Act.
BACKGROUND: Industry manufacturers are required by the Sunshine Act to disclose payments to physicians. These data recently became publicly available, but some manufacturers prereleased their data since 2009. We tested the hypotheses that there would be discrepancies between manufacturers' and physicians' disclosures. METHODS: The financial disclosures by authors of all 39 American College of Cardiology and American Heart Association guidelines between 2009 and 2012 were matched to the public disclosures of 15 pharmaceutical companies during that same period. Duplicate authors across guidelines were assessed independently. Per the guidelines, payments <$10,000 are modest and ≥$10,000 are significant. Agreement was determined using a κ statistic; Fisher's exact and Mann-Whitney tests were used to detect statistical significance. RESULTS: The overall agreement between author and company disclosure was poor (κ = 0.238). There was a significant difference in error rates of disclosure among companies and authors (P = .019). Of disclosures by authors, companies failed to match them with an error rate of 71.6%. Of disclosures by companies, authors failed to match them with an error rate of 54.7%. CONCLUSIONS: Our analysis shows a concerning level of disagreement between guideline authors' and pharmaceutical companies' disclosures. Without ability for physicians to challenge reports, it is unclear whether these discrepancies reflect undisclosed relationships with industry or errors in reporting, and caution should be advised in interpretation of data from the Sunshine Act.
Authors: Jennifer C Thompson; Katherine A Volpe; Lindsay K Bridgewater; Fares Qeadan; Gena C Dunivan; Yuko M Komesu; Sara B Cichowski; Peter C Jeppson; Rebecca G Rogers Journal: Am J Obstet Gynecol Date: 2016-06-16 Impact factor: 8.661
Authors: Sahar Tabatabavakili; Rishad Khan; Michael A Scaffidi; Nikko Gimpaya; David Lightfoot; Samir C Grover Journal: Mayo Clin Proc Innov Qual Outcomes Date: 2021-01-19
Authors: Anna R Gagliardi; Pascale Lehoux; Ariel Ducey; Anthony Easty; Sue Ross; Chaim Bell; Patricia Trbovich; David R Urbach Journal: PLoS One Date: 2017-03-30 Impact factor: 3.240