| Study characteristics |
| Methods |
To explore the association between authors’ financial conflicts of interest and their position on the association between rosiglitazone in patients with diabetes and cardiovascular events |
| Data |
5 clinical guidelines, 91 opinion pieces (letters, editorials, commentaries), and 84 narrative reviews |
| Comparisons |
Documents with financial conflicts of interest (defined as at least one author with funding of the document, employment, consultancy, advisory board membership, speaker or lecture feeds, travel grants, stock ownership or honoraria from pharmaceutical companies) and documents without financial conflicts of interest |
| Outcomes |
Recommendations (favourable recommendations defined as recommending the use of rosiglitazone) |
| Funding source |
No funding was received for the study |
| Declaration of conflicts of interest |
The authors disclosed no conflicts of interest related to any for‐profit company |
| Notes |
|
| Risk of bias |
| Item |
Authors' judgement |
Support for judgement |
| Adequate document inclusion process |
Yes |
Two authors independently assessed documents for inclusion (according to personal correspondence with corresponding author) |
| Adequate coding of conflicts of interest |
No |
Only one author coded each document |
| Adequate coding of recommendations |
Yes |
Two review authors independently coded the recommendations of each document |
| Adequate dealing with confounding |
Yes |
Compared documents on the same drug used for the same disease |
