Do drug firm ties affect researchers’ reporting of study results?