Statistics from Altmetric.com
Statistical methods are important. We have all seen the impact of shifting from p values to confidence intervals and from relative (RRR) to absolute measures (ARR, NNT) of effect, and how these can influence the interpretation and implementation of study results. A continuing controversy is over the use of multiple outcomes. In this issue, we thought the results of the study by Crowther et al important enough to be one of the 20 articles we chose from the 100,000 published in the past 2 months. But we’ve accompanied this by an editorial on the interpretation issues. We would welcome your comments on this difficult issue.
If you enjoy discussion about EBM topics, then you might wish to sign on to the Evidence-Based Health email list at: www.jiscmail.ac.uk/lists/EVIDENCE-BASED-HEALTH.html
In addition to alerts about forthcoming courses related to EBM, the list has a lively discussion of topics in evidence-based health care. Some recent threads have included an explanation and history of Simpson’s paradox, outcomes research, the impact of dishonest reporting of research on EBM, and of course, that perennial discussion about the relative merits of RCTs and observational studies. In the course of that thread Badri Badrinath pointed out the fascinating article by John Ioannidis (Contradicted and initially stronger effects in highly cited clinical research.
), in which the law of initial results ("first results are always spectacular, and subsequent ones are mediocre") was tested using the results of subsequent research. The author found that "most of the observational studies had been contraindicated, or subsequent research shown substantially smaller effects, but most randomised studies had results that had not been challenged." Happy reading,
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.