Setting the results of individual studies in context
- Iain Chalmers, MSc (Soc Med)
Felicitations on the birth of Evidence-Based Nursing (EBN) and best wishes for the future. I was particularly glad to see that you have made a start on working out how to select reliable studies reporting qualitative data.
The first issue of EBN prompts me to write with a suggestion which I hope you will consider implementing as the journal matures. I think it is potentially misleading to concentrate on abstracting individual studies (as EBN has begun to) unless either the authors or the commentators have made a systematic effort to set the results of these studies within the context of other relevant studies. For example, the study on the effects of sucrose on crying in newborn babies which you selected for abstracting is 1 of at least 5 relevant controlled trials, yet these other trials are not mentioned in the commentary. In fact, the relevant trials on this topic have been assessed in a good systematic review which must have been published at about the same time as the single study you selected for presentation.1 In my view, the systematic review provides evidence of much greater importance to your readers than the outcome of the particular study you selected (in which the phrase `Ross Special Formula' is mentioned three times!).
Unless the results of the individual studies you select for presentation in EBN are systematically set in proper context, either by the authors or by your commentators, your readers are likely to be misled—by the play of chance among similar studies if by nothing else.
Thank you for welcoming the first issue of EBN and for taking the time to give us this thought provoking feedback. Your letter raises important issues that we must grapple with as the journal develops. We agree that it is potentially dangerous to present the findings of isolated studies when there is an existing systematic review.
We endeavour to bring relevant, existing systematic reviews to the attention of each commentator, who is briefed to set the new research in the context of existing knowledge. Unfortunately, in this case the meta-analysis by Stevens et al became available after we had gone to press and the commentator was unable to refer to it in her commentary. Nevertheless, your letter is a useful reminder of the importance of setting individual study results in context. Evidence-Based Nursing places high priority on abstracting recent systematic reviews of relevance to nursing, and we hope to increase our coverage as more nursing related systematic reviews become available.