What's new in the other general journalsBMJ 2006; 332 doi: https://doi.org/10.1136/bmj.332.7543.713 (Published 23 March 2006) Cite this as: BMJ 2006;332:713
- Alison Tonks, associate editor (firstname.lastname@example.org)
Results often change between scientific meetings and full publication
Most big scientific meetings have sessions reserved for important late breaking trials. Between 1999 and 2002, 86 trials made it into these sessions at meetings of the American College of Cardiology. Unsurprisingly they were bigger, better, and more likely to be subsequently published than trials presented in other sessions, a recent study has found. Despite their high quality and impact, reports of late breaking trials were just as likely as other trials to change between the meeting stage and full publication up to three years later. Overall, 41% of trials reported different effect sizes, and a quarter reported different sample sizes in the initial report and the full paper. In one trial in seven, the statistical significance of the effect also changed. Effect sizes changed by more than one standard deviation, on average.
The authors don't explore in detail why discrepancies are so common, but at least some of the sample sizes changed because preliminary results presented at the meetings were later extended. The authors do question, however, whether early reports of important papers should be included in the evidence base for treatments, since so many of them seem unstable.
Traditional treatment for croup doesn't work
Doctors have been treating croup with humidified air for well over 100 years. That's probably long enough, say researchers from Canada, after their carefully controlled and blinded trial failed to find any evidence of benefit. Children given 40% oxygen at 100% humidity did no better than children who were effectively given enriched room air to breathe. One control group had the standard humidification from flexible tubing directed towards the face by a parent. The other control group had 40% oxygen at 40% humidity via a face mask. The researchers did not include an untreated group because humidification is the standard treatment for croup.
Although the trial was small …