RSS Feeds
-
Suggestions for the authors
Submit responseI read with great interest Shojania and colleagues article [1] on how quickly systematic reviews go out of date. It is of note that they first applied survival analysis methodology to adequately assess the median survival time (5.5 years) for 100 highly qualified systematic reviews. Although they have set up reasonable criteria for both quantitative and qualitative signals for update, their definition regarding pivotal trials, which is trials with a sample size of at least 3 times larger than that of the previous largest trial, or published in one of five highest-impact general medical journals, needs further justification. Theoretically, a trial with twice the sample size of the previous largest trial that somehow failed to be published in those five journals (non-pivotal by their definition) would not contribute to the updating process.
In addition, unlike in the article, a broader definition of updating systematic reviews is needed. It should include not only invalidating information or novel clinical caveats, but also backing evidence that further consolidates the current practice guidelines, or clinical recommendations. Unless systematic reviews encompass recent trials worthy of being informed to the medical society, they can go out of date even without changes in major findings of the original article.
Reference: 1. Shojania KG, Sampson M, Ansari MT, Ji J, Doucette S, Moher D. How quickly do systematic reviews go out of date? A survival analysis. Ann Intern Med. 2007; 147:273-4.
Conflict of Interest:
None declared
-
Rejection of Study Claims
Submit responseIn their article on the longevity of the recommendations from systematic reviews Shojania et al. properly identify the critical need to regulary update reviews. However, they do not provide an answer to an equally compelling question. How many of the recommendations of systematic reviews are subsequently rejected by newer studies and how quickly does this occur? Specific rejection of findings – no effect where the review claimed one or an effect where the review explicitly rejected one is what I am seeking. Extension of efficacy to populations not studied before with adequate power does not constitute rejection. I think we would all be benefited by the authors providing such a table perhaps with some detail of the nature of the rejection as categories. As a consumer of journals, reviews, cd lecture series and mksap, I am often struck by conflicting claims and it would be useful to have a quantification of this phenomenon in the subset studied by the authors.
(1)Kaveh G. Shojania, Margaret Sampson, Mohammed T. Ansari, Jun Ji, Steve Doucette, and David Moher How Quickly Do Systematic Reviews Go Out of Date? A Survival Analysis Ann Intern Med 2007; 147: 224-233
Conflict of Interest:
None declared