An investigation into drug-company influence on research studies they've funded revealed that the researchers were often encouraged to arrive at a 'favourable conclusion' even when the results and small print didn't warrant one. Interestingly, the drug company rarely asked for the study itself to be changed, which might puzzle the more observant reader.
Another common problem is 'ghosting', where an eminent professor agrees to put his name to the work of some unknown researcher. This adds credibility to the findings (especially if, as expected, they're favourable). It's unusual for the professor to ever have read the study.
Finally, as we've seen before on these pages, the suppression of papers is not uncommon if the findings are so unfavourable that even the gloss of a 'doctored' (no pun intended) conclusion cannot salvage it. Studies that discover that a drug is actually killing patients tend never to see the light of day.
These practices have become so rife that 470 editors and publishers of scientific and medical journals have formed a group dedicated to improving the scientific validity of submitted papers.
However, then there's the tricky issue of the millions of dollars of drug-company advertising in their magazines to consider, of course . . . (J Am Med Assoc, 2005; 294: 2287-8).