Commentary: but what should journals actually do to keep industry sponsored research unbiased?

A timely paper from Robert Steinbrook and Jerome P Kassirer (doi:10.1136/bmj.c5391) takes the position that journals should increase the rigour required to address bias in data offered for publication from industry sponsored trials. In particular (1) the definition (currently ambiguous) of “full access to data” for the authors should be cleared up; (2) one author in the team should be truly independent of any pecuniary interest in the product and act as guarantor; (3) data must be available to editors in case of future concern, (something hopefully rarely exercised); and finally (4) there would be consequences after publication if data were found to have been …