Examples, mechanisms and management of conflicts of interest in randomized clinical trials: qualitative interview study

Article type
Authors
Østengaard L1, Lundh A2, Ahmed Ibrahim Sheekh Abdi S3, Hassan Adan Gelle M3, Stewart L4, Boutron I5, Hróbjartsson A6
1Centre for Evidence-Based Medicine Odense (CEBMO), Odense Patient data Exploratory Network (OPEN), Department of Clinical Research, University of Southern Denmark, University Library of Southern Denmark, Odense
2Centre for Evidence-Based Medicine Odense (CEBMO), Odense Patient data Exploratory Network (OPEN), Department of Clinical Research, University of Southern Denmark, Department of Infectious Diseases, Hvidovre Hospital, Hvidovre
3Centre for Evidence-Based Medicine Odense (CEBMO)
4Centre for Reviews and Dissemination, University of York
5Université Paris Descartes, Cochrane France, Equipe METHODS, Centre de Recherche Épidémiologie et Statistique Sorbonne Paris Cité (CRESS-UMR1153) Inserm, Paris
6Centre for Evidence-Based Medicine Odense (CEBMO), Odense Patient data Exploratory Network (OPEN), Department of Clinical Research, University of Southern Denmark, Odense
Abstract
Background: conflicts of interest may influence how clinical trials are designed, conducted, analysed and reported. Published reports have described flaws in several examples of industry-funded clinical trials. However, our knowledge is incomplete regarding by which mechanisms conflicts of interest may influence a clinical trial and what management is done in trials in order to minimize the possible influence from conflicts of interest.

Objectives: to characterize and analyze trialists’ experience with how conflicts of interest have influenced clinical trials and which procedural elements they have implemented to minimize such influence in clinical trials.

Methods: this project is a qualitative interview study with experienced trialists. We pragmatically identified the possible interview candidates through snowball sampling. We targeted our search for participants worldwide and wanted to ensure a broad representation by geography, clinical areas, professional disciplines and experience with both academia and industry. We developed an interview guide, which we pilot tested prior to the interviews. We stopped including new participants when no new major themes were emerging. The interviews were recorded and transcribed verbatim. We used systematic text condensation to analyze the data for a thematic cross-case analysis in NVivo 12 Pro.

Results: we conducted 20 interviews in the period December 2017 to July 2018. Eight out of 20 respondents stated that they had participated in a clinical trial where someone with conflicts of interest had tried to unduly influence the trial.

Four major themes emerged from the interviews:
1) definition of conflicts of interest;
2) impact of conflicts of interest;
3) reporting of conflicts of interest; and
4) management of conflicts of interest.

Additional elaboration of themes awaits further analysis, which will be ready for presentation at the 26th Cochrane Colloquium in Santiago, Chile.

Conclusions: final conclusion awaits further analysis. The findings from this study will provide an insight into how conflicts of interest can influence clinical trials and which procedural elements can be implemented to minimize the influence of conflicts of interest.

Patient or healthcare consumer involvement: none