The study by Greene et al.  is a useful complement to our 2007 study, in which we observed a larger effect for NRT in industry-sponsored trials than in independent trials . We offered two explanations for this result. Industry-sponsored trials may have more resources, leading to higher compliance with treatment recommendations and, thus, to larger treatment effects. There may also have been a publication bias among industry-supported trials, as a funnel plot suggested that several industry-sponsored trials with small sample sizes never reached publication. Greene et al. found support for a third hypothesis: lower success rates in placebo groups in industry-sponsored trials, and they offer several explanations for this result. This work is a useful addition to ours and underlines the necessity to disclose conflicts of interests and sources of funding, in order to maintain the public's confidence in medical research.
Declaration of interests
The Faculty of Medicine of the University of Geneva received funding from manufacturers of nicotine replacement products to develop internet-based smoking cessation programs, under the supervision of JFE, who also acted as an adviser to manufacturers of smoking cessation medications, but he declares no conflict of interest after 2007. JS has acted as an adviser to several organizations with an interest in smoking cessation, including manufacturers of nicotine replacement products. He has also conducted a number of UKMRC-funded clinical trails that also received support from the manufacturers of nicotine replacement products. JS declares no conflict on interest after 2005.