The Lancet  
Dec 10, 2011  Volume 378  Number 9808  p1975 – 2048  e19 – 21
http://www.thelancet.com/journals/lancet/issue/current

Editorial
What constitutes full access to data in industry-funded trials?
The Lancet

Preview
Open any medical journal and you are likely to find significant results outnumbering those that are non-significant. Novel, positive findings evoke greater interest than do confirmatory or “negative” trials, although the latter are equally as important. Such publication bias at a journal level has long been judged problematic. But might selective data reporting at a study level be even more sinister? The answer is yes.