Potential Bias in Image-Guided Procedure Research: A Retrospective Analysis of Disclosed Conflicts of Interest and Open Payment Records

Accuracy of Financial Disclosures in Radiology Journals

PURPOSE

The accuracy and completeness of self-disclosures of the value of industry payments by authors publishing in radiology journals are not well known. The aim of this study was to assess the accuracy of financial disclosures by US authors in five prominent radiology journals.

METHODS

Financial disclosures provided by US-based authors in five prominent radiology journals from original research and review articles published in 2021 were reviewed. For each author, payment reports were extracted from the Open Payments Database (OPD) in the previous 36 months related to general, research, and ownership payments. Each author was analyzed individually to determine if the reported disclosures matched results from the OPD.

RESULTS

A total of 4,076 authorships, including 3,406 unique authors, were selected from 643 articles across the five journals; 1,388 (1,032 unique authors) received industry payments within the previous 36 months, with a median total amount received per authorship of $6,650 (interquartile range, $355-$87,725). Sixty-one authors (4.4%) disclosed all industry relationships, 205 (14.8%) disclosed some of the OPD-reported relationships, and 1,122 (80.8%) failed to disclose any relationships. Undisclosed payments totaled $186,578,350, representing 67.2% of all payments. Radiology had the highest proportion of authorships disclosing some or all OPD-reported relationships (32.3%), compared with the Journal of Vascular and Interventional Radiology (18.2%), the American Journal of Neuroradiology (17.3%), JACR (13.1%), and the American Journal of Roentgenology (10.3%).

CONCLUSIONS

Financial relationships with industry are common among US physician authors in prominent radiology journals, and nondisclosure rates are high.

Impact of the IR/DR Primary Specialty Certification on Scientific Research in the Field

Research is essential to the development, maintenance, and success of interventional radiology (IR) as a primary specialty. IR physicians must embrace a more sophisticated research approach toward the procedures we perform and the diseases we treat to accompany our elevated status. Identifying impactful research needs and directing a comprehensive research portfolio will anchor our position as an essential specialty within a thriving health care system. Doing so will ensure that this historic opportunity is maximized for the benefit of IR physicians, IR trainees, and, ultimately, IR patients.

The undisclosed disclosures: The dollar-outcome relationship in resuscitative endovascular balloon occlusion of the aorta

BACKGROUND

Despite its rapid evolution, resuscitative endovascular balloon occlusion of the aorta (REBOA) remains a controversial intervention that continues to generate active research. Proper conflict of interest (COI) disclosure helps to ensure that research is conducted objectively, without bias. We aimed to identify the accuracy of COI disclosures in REBOA research.

METHODS

Literature search was performed using the keyword "REBOA" on PubMed. Studies on REBOA with at least one American author published between 2017 and 2022 were identified. The Centers for Medicare and Medicaid Services Open Payments database was used to extract information regarding payments to the authors from the industry. This was compared with the COI section reported in the manuscripts. Conflict of interest disclosure was defined as inaccurate if the authors failed to disclose any amount of money received from the industry. Descriptive statistics were performed.

RESULTS

We reviewed a total of 524 articles, of which 288 articles met the inclusion criteria. At least one author received payments in 57% (165) of the articles. Overall, 59 authors had a history of payment from the industry. Conflict of interest disclosure was inaccurate in 88% (145) of the articles where the authors received payment.

CONCLUSION

Conflict of interest reports are highly inaccurate in REBOA studies. There needs to be standardization of reporting of conflicts of interest to avoid potential bias.

LEVEL OF EVIDENCE

Prognostic and Epidemiological; Level IV.