Monday, 24 March 2014

Trial by Author

Roger Watson, Editor-in-Chief

Periodically, authors scrutinise what is published in journals and then publish it. I welcome this and, with respect to h-indices (Thompson & Watson 2010), open access (Watson et al. 2012), peer review (Watson 2012), authorship (Hayter et al. 2013) and citations (Hunt et al. 2013, Watson et al. 2013) I have done some of this myself. Publishing about publishing is increasingly common and, while some may see this as futile self-absorption, it is important in exposing and improving standards.

A recent JAN paper by Chiavetta et al. (2014) considers whether there is a difference in methodological quality between positive and negative published clinical trials. They conclude that there is and that negative trials tend to do better when scored using the Jadad scale which looks at randomisation, blinding and how all patients are accounted for. The years studied were 2010-2012 and, since the journals they analysed include JAN, my attention was drawn to this paper by our Managing Editor.

Ideally, there should be no difference in the methodological quality of positive or negative clinical trials and none of us - without appropriate qualification or description (e.g. pragmatic, single-blind, quasi-experimental) - should be publishing items labelled clinical trials that are not described fully. For the details of the disparate points between trials I urge you to read the paper; of more interest to me as Editor-in-Chief of JAN is ‘why?’. Why was there a difference and, of course, what can we do about it?

The authors speculate that it is still hard to publish the results of negative clinical trials; this should not be true but it may be. At JAN we do not have a policy of not publishing negative trials and I hope we convey that message. Nevertheless, as I travel the world and give writing workshops I am frequently asked about our policies on negative trials. Clearly, people do find it hard to publish these and whatever efforts we are making to dispel this, we are not making them strongly enough.

The argument by the authors, therefore, runs as follows: it is harder to publish negative trials and they come under greater scrutiny than positive trials; thus the standard of their publication is driven up. However, editors and reviewers like positive trials and they receive less scrutiny and ‘slip the net’ of the reviewing and editing process. If this is true then we need to work harder to obviate this as it is surely a contributory factor to publication bias and this is bad because poor positive results are published and exaggerated. This leads to misuse of resources and could even endanger patients.

The solution proposed by Chiavetta et al. (2104) includes the use of the CONSORT guidelines which we do recommend in JAN there may be a case for emphasising this in our guidelines and strengthening its implementation by authors, reviewers and editors.

References

Chiavetta N, Martins ARS, Henriques ICR, Frengi F (2014) Differences in methodological quality between positive and negative published clinical trials Journal of Advanced Nursing doi: 10.1111/jan.12380

Hayter M, Noyes J, Perry L, Pickler R, Roe B, Watson R (2013) Who writes, whose rights, and who’s right? Journal of Advanced Nursing 62, 2599-2601

Hunt GE, Jackson D, Watson R, Cleary M (2013) A citation analysis of nurse education using various bibliometric indicators Journal of Advanced Nursing 62, 1411-1445

Thompson DR, Watson R (2010) h-indices andthe performance of nursing professors in the UK Journal of Clinical Nursing 19, 2975-2958

Watson R (2012) Peer review under the spotlight in the UK Journal of Advanced Nursing 68, 718-720

Watson R Cleary M, Jackson D, Hunt GE (2012) Open access and online publishing: a new frontier in nursing? Journal of Advanced Nursing 68, 1905-1908

Watson R, Cleary M, Hunt GE (2013) What gets highly cited in JAN? Can editors pick articles which will contribute to a journal’simpact factor? Journal of Advanced Nursing doi:10.1111/jan.12261

No comments:

Post a Comment