On 2015 Nov 03, John Tucker commented:

The authors find evidence in support of the unsurprising conclusion that pharmaceutical employees are loathe to criticize their own company's products. I don't think anyone will contest that. They also find that academic authors who have accepted research funding from industry currently or in the past are less likely than those who have not to make critical remarks in the paper abstract, and conclude that this represents the influence of conflict-of-interest.

While the first conclusion is both clear and expected, the second involves an implicit and unproven assumption that those who eschew all contact with industry represent an absolute standard of objectivity against which others can be measured. While one can readily point a finger at research support as a conflict, the same reasoning requires that we ask why others have turned research money down. Is it purely a desire for both the actuality and appearance of complete objectivity? Does it stem from a pre-existing distrust and dislike of for-profit drug development? All we really know is that there are likely systematic differences between those who work with industry to develop drugs and those who keep industry at arms length.

One way to address this question might be to take the list of authors of meta analyses included in this study and assess the extent to which positive/negative assessments of individual drugs correlates with public expressions of positive/negative opinion about the pharmaceutical industry as a whole. I suspect that the observed correlation is high, and that when reading any meta analysis, one should be aware that the subjective and non-transparent processes of study selection and data reduction makes most meta analyses little more than detailed editorials.

On 2015 Nov 03, John Tucker commented:

The authors find evidence in support of the unsurprising conclusion that pharmaceutical employees are loathe to criticize their own company's products. I don't think anyone will contest that. They also find that academic authors who have accepted research funding from industry currently or in the past are less likely than those who have not to make critical remarks in the paper abstract, and conclude that this represents the influence of conflict-of-interest.

While the first conclusion is both clear and expected, the second involves an implicit and unproven assumption that those who eschew all contact with industry represent an absolute standard of objectivity against which others can be measured. While one can readily point a finger at research support as a conflict, the same reasoning requires that we ask why others have turned research money down. Is it purely a desire for both the actuality and appearance of complete objectivity? Does it stem from a pre-existing distrust and dislike of for-profit drug development? All we really know is that there are likely systematic differences between those who work with industry to develop drugs and those who keep industry at arms length.

One way to address this question might be to take the list of authors of meta analyses included in this study and assess the extent to which positive/negative assessments of individual drugs correlates with public expressions of positive/negative opinion about the pharmaceutical industry as a whole. I suspect that the observed correlation is high, and that when reading any meta analysis, one should be aware that the subjective and non-transparent processes of study selection and data reduction makes most meta analyses little more than detailed editorials.