Reforming Research and Publication Methods

In our previous article on open data (http://cwauthors.com/article/opendata), we discussed that data sharing allows other researchers to evaluate the work of authors and facilitates the methodologies for carrying out follow-up studies to validate whether the results in the previously published work are reliable and reproducible.

Large quantities of research are being published all the time; with the constant developments taking place that confirm or negate previous studies, sometimes contradictory information is reported in popular media outlets such as printed and online newspapers. This contradictory information can lead to a mistrust of health information among the general public, and sometimes even within the scientific community.

Arguably, the transparency associated with open data would facilitate the eventual publication of reproducible research results, thereby increasing the overall value of published research.

To the same ends, another idea has been proposed and described by Nature, whereby more extensive research is carried out as part of a “preclinical trial” to eliminate weak hypotheses before they reach the later stages of clinical trials.[1] These confirmatory studies would be required to provide substantial justification for further research and clinical research.

Publishing negative results

The Journal of Negative Results in BioMedicine (JNRBM) represents another means of highlighting inadequate or incomplete research by publishing “results that challenge current models, tenets or dogmas”.[2] In doing so, it provides “responsible and balanced information” in an effort to counter the “insufficient evidence regarding negative data” that is “frequently” provided in “articles published in traditional journals”.[3]

Confirmatory studies

While JNRBM seeks to publish negative results to protect against ill-informed reliance on research, Nature proposes a new type of research paper to encourage researchers to strive for reliable and reproducible results rather than simply acceptance and publication in high-impact journals. According to Nature, many of the current practices in clinical trials would be performed at the preclinical trial stage and the new research papers would include exploratory and confirmatory studies.[4]

Research impact

Preclinical trials as outlined above “would allow much more flexibility in earlier hypothesis-generating experiments”, and publishing them alongside the confirmatory research would result in “fewer high-profile papers hailing new therapeutic strategies, but much more confidence in their conclusions”.[5] This is because new therapies would not be published before their value has been “confirmed” during the preclinical trials.

This practice would also lead to a higher publication rate of negative results, as encouraged by JNRBM: negative and positive results alike indicate the status of a hypothesis.

At the moment, pharmaceutical companies and other organizations require increasingly stricter guidelines concerning the conduct of preclinical research, which arguably restricts the efficiency and creativity of researchers.[6] Under Nature’s proposed methods, researchers could be more creative with their hypotheses at exploratory stages, which would create greater potential for innovative findings. Any hypothesis could be tested but, before publication and clinical trials, rigorous and independent confirmatory trials would be held.

Impact on clinical trials

It is argued that the rigor of preclinical trials would mean that clinical trials would be safer for human patients. Furthermore, although the extensive preclinical trials are likely to require up to six times the number of animal test subjects, the scope of these confirmatory tests would mean that fewer follow-up studies (such as those published in JNRBM) would be required, thus saving animal and monetary resources down the line. [7]

Funding

A complete reshuffle in the way that research is conducted has been proposed:[8] it may be that independent laboratories are founded with researchers specializing in confirmatory research. Perhaps researchers would collaborate according to speciality, and companies may redistribute funds, with more at the exploratory stages and less required as the research narrows.

Conclusion

As suggested by JNRBM, negative results are important indicators of the reliability of a study which must be published. The aim of Nature’s proposed research method and publication process is to eliminate poorly conducted research and reduce its integration into published articles. Indeed, as the confirmatory research would require almost infallible confirmation of results, this is likely to reduce the number of published studies that support new treatment methods or medication; instead, the publication of papers presenting a more in-depth analysis and more reliable results is likely to increase.

Additional help and support