On 2016 Jun 02, John Tucker commented:

The authors of this paper undertook a survey of published results for trials conducted between January 2000 and December 2006. The results of this survey were that industry-only sponsored studies were more likely to obtain positive results than those partially funded by industry, and that these were in turn more likely to report positive results than those performed without industry funding. The results of this study have been widely cited as evidence that the results of industry funded studies are unreliable and tainted by bias.

A fundamental difficulty with this type of analysis is the underlying assumption that the trials performed by industry and those performed without industry funding are substantially comparable. While the authors do not provide sufficient detail regarding their search criteria to allow perfect recapitulation, a search of the clinicaltrials.gov database with the term "hypercholesterolemia" turns up the following trials.

*93 funded by industry alone, including 39 trials of statins and 32 trials of ezetimibe

*7 funded jointly by industry and some other entity, including trials of orange juice, a high fat diet, and glucosamine as interventions

*7 funded entirely by non-industry entities, including studies of garlic (2x), flaxseed (2x), a high protein diet (2x), and plant sterol-enriched tea as interventions.

Given the vast gulf between the types of interventions being examined and their a priori likelihood of efficacy, the results of studies such as this one should be interpreted with caution.

On 2016 Jun 02, John Tucker commented:

The authors of this paper undertook a survey of published results for trials conducted between January 2000 and December 2006. The results of this survey were that industry-only sponsored studies were more likely to obtain positive results than those partially funded by industry, and that these were in turn more likely to report positive results than those performed without industry funding. The results of this study have been widely cited as evidence that the results of industry funded studies are unreliable and tainted by bias.

A fundamental difficulty with this type of analysis is the underlying assumption that the trials performed by industry and those performed without industry funding are substantially comparable. While the authors do not provide sufficient detail regarding their search criteria to allow perfect recapitulation, a search of the clinicaltrials.gov database with the term "hypercholesterolemia" turns up the following trials.

*93 funded by industry alone, including 39 trials of statins and 32 trials of ezetimibe

*7 funded jointly by industry and some other entity, including trials of orange juice, a high fat diet, and glucosamine as interventions

*7 funded entirely by non-industry entities, including studies of garlic (2x), flaxseed (2x), a high protein diet (2x), and plant sterol-enriched tea as interventions.

Given the vast gulf between the types of interventions being examined and their a priori likelihood of efficacy, the results of studies such as this one should be interpreted with caution.