Biases which begin at the end

Scientific inquiry and proper dissemination of research information are quintessential for the development and progress of scientific knowledge. However, bias, in its various forms, has the capability to disrupt the integrity of research at every level and significantly skew scientific research, affecting its interpretation and reliability. Every research must logically end in its sharing with everyone (publication); while the biases which afflict the conduction of research are more commonly discussed and acknowledged, the biases during publication are just as prevalent and equally important, which come into play at the end of the research.

One of the most pervasive forms of bias in research publication is publication bias, whereby studies with statistically significant results are more likely to be published than those with null or nonsignificant findings.[1,2] In general, scientific community tends to favor studies with positive results, which may demotivate a researcher to report a study with negative outcomes, thus creating a gap between the published and unpublished data. When a result contrary to prior publications and scientific literature is obtained by the researchers, it is assumed that the study may be flawed, or there may be apprehension of loss of funding for the study and the investigators may not be motivated to publish it. The competition for publication may incite a researcher to dredge data[3] and somehow report a positive result since it is more likely to find acceptability with editors and reviewers.[4] A positive publication may get more citation, leading to a citation bias. For editors, it is the struggle to maintain citation indexes and impact factors on which the financial viability of the journal depends, which makes it more attractive to publish positive findings.[5,6]

Publication bias alters scientific literature, highlighting the positive results and underemphasizing negative correlations, thereby distorting our understanding of the true cause and effect of the phenomenon under investigation. It is more evident in systematic reviews and meta-analyses which are considered as level 1 category evidence-based clinical practice.[7] These are based on pooling of data from published clinical trials, hence publication bias (absence of studies with negative result) will skew the final analysis in favour of positive outcomes. The classical example of publication bias has been reported in the field of psychology where antipsychotic drugs have been proven to be of variable efficacy after inclusion of unpublished data.[8,9] Hence, we need to reconsider the amount of credibility given to these publications unless they include studies with negative results. In Ophthalmology as well, it has been reported that the studies with positive results have better odds of publication in journals with high impact factors.[10]

The publication bias could impact the decision of policy makers and guide/misguide further research in the field, and when this bias percolates into clinical practice, it could have major health consequences for the population and directly or indirectly influence the other stakeholders. Research publications need to embody high-quality research, and publishers need to maintain neutrality vis-a-vis the direction or magnitude of results. Researchers on their part too need to be aware of and sensitive to the same effects and need not shy away from submitting research of true value.

Apart from publication bias, editorial bias may arise when journals’ publication is affected by the type, topic, or geographical location of research, leading to disparities in the visibility and recognition of diverse scientific contributions. Similarly, there is possibility of bias during the peer review process; the lack of interest and ennui in overburdened reviewers unwilling to rigorously evaluate the manuscript under consideration, undue preference for favorite topics or vice versa, and inability to keep up to date are potential sources of bias during the peer review.[11,12] Institution of diversified editorial boards and reviewer panels may help to mitigate editorial bias. Journals should adopt robust peer review processes that emphasize methodological rigor, scientific merit, and scholarly significance, while minimizing subjective biases and prejudices.

Another form of bias that can compromise the integrity of research publication is conflict of interest in researchers, publishers, editors, and reviewers.[13] In addition to financial interests, professional or personal interests may unduly influence the conduct, interpretation, or reporting of research findings. To address conflicts of interest, journals should implement stringent disclosure requirements, mandating both researchers and reviewers to be transparent about any potential conflicts that could influence their research or review. Such disclosures should be carefully scrutinized for their potential impact on research findings.

In conclusion, bias in the publication of research poses significant challenges to the integrity and reliability of scientific knowledge. Acknowledgment of the presence of bias and understanding the nature of bias should be followed by implementation of proactive measures to address its root causes and mitigate its impact. The role of impact factor in the financial viability of the journal needs to be looked at through broader lens in order to make publication a fair process. Training programs and awareness initiatives to sensitize researchers, editors, and reviewers about nature and influence of bias would go a long way in ensuring credibility and accountability within the scientific community. As custodians of science, publishers, editors, reviewers, and researchers must be committed to foster a culture of fairness and excellence in scientific publication, ensuring that the pursuit of knowledge remains guided by the timeless principles of truth, objectivity, and intellectual integrity.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.