Table 1.

Examples of the impact of undisclosed conflicts of interest on clinical evidence, public opinion, and clinical decision-making

Interventions	The potential impact of undisclosed conflicts of interest
Rosiglitazone	Following a meta-analysis showing an association between rosiglitazone and cardiovascular risk, articles authored by researchers with conflicts of interest were more likely to uphold the safety of the drug [11]. Among the articles with identified conflicts of interest, 23 % did not disclose them. Rosiglitazone was withdrawn from the market for safety reasons in several countries but remains available in the USA.
Alteplase	Alteplase was strongly recommended for use in acute stroke in clinical guidelines despite resistance from emergency physicians concerned about intracerebral hemorrhage [84]. Seven of eight panelists developing the guidelines had potential conflicts of interest (indirect financial ties to the manufacturer of alteplase), but only three of the panelists disclosed these conflicts [85]. After the conflicts of interest were revealed, the American Heart Foundation withdrew statements that the intervention could save lives [13].
Risperidone	While failing to completely disclose financial relationships with the manufacturer of risperidone, an influential researcher was instrumental in expanding the diagnosis criteria for bipolar disorder in children and conducted a number of pediatric clinical trials demonstrating the benefit of the drug in children [86]. A congressional investigation later found him guilty of violating federal and university regulations and conflicts of interest policies.
Calcium-channel antagonists	A survey study found that authors’ published positions on the safety of calcium channel antagonists were more likely to be favorable to the drug class if they responded that they had a financial conflict of interest (63 % of authors reported a financial conflict of interest in the survey) [12]. However, only 2 of the 70 articles authored by the respondents included disclosures.
Measles, mumps, rubella (MMR) vaccine	A study linking the MMR vaccine to autism was eventually retracted after it was discovered that an author failed to disclose how he stood to gain financially by discrediting the vaccine [87, 88]. The impact on vaccine decision-making persists even a decade later, with surveys showing that more than one in five people believe that vaccines cause autism [89].
Neuraminidase inhibitors	Academics who were interviewed in newspaper articles covering the 2009 H1N1 pandemic were more likely to overestimate the risk of the pandemic or promote the use of neuraminidase inhibitors if they had conflicts of interest [28]. Only 3 of 425 newspaper articles noted the academics’ conflicts of interest.