.png.aspx "Untitled-design-(16).png")
What is research integrity?
Research integrity is defined as conducting research in a manner that enables the audience to have the utmost confidence and trust in the findings and data presented. Research integrity is essential for conducting scientific research. As scientists and researchers, we must advocate for and implement the values of honesty, accuracy, efficiency and objectivity, and ensure that the ways we conduct and analyse research meet research integrity requirements.
The UK Research Integrity Office (UKRIO) is an independent charity that supports researchers and academic institutes by providing guidance for research publication and supporting good research practice. In 2012, the UKRIO helped to develop the UK Concordat to Support Research Integrity - the national research integrity policy statement signed by approximately 140 universities. The concordat outlines the key elements of research integrity, which include honesty, rigour, transparency, open communication, the care and respect of all participants, and accountability.
It is important to recognise that failings in research and scientific integrity can occur either deliberately (through scientific misconduct) or accidentally (through human error). Reasons for research integrity failings can include academic pressures, lack of knowledge, peer and supervisor pressure, and poor time and data management.
In this article, we discuss the importance of research integrity, the role that the academic publishing process plays in setting and upholding standards, how the COVID-19 pandemic has increased the importance of these conversations, and how we can strive for better overall standards of research integrity.
Why is research integrity important?
Upholding research and scientific integrity are of paramount importance for many reasons. The first of these is trust. Research progresses by the exchange of information between researchers. This exchange helps researchers to develop ideas, improve existing techniques and protocols, refine study designs, and produce quality data that can shape our understanding of topics.
To build on scientific knowledge, scientists need to be able to trust the work carried out in the published studies they draw from, particularly as we often design new research based on the findings of existing studies. In this process, researchers rely on peer review to address any errors in experimental design and the resulting data. This review process is important because if new studies are built upon inaccurate information, it can result in the wasting of time, resources, and funding.
A significant proportion of research relies on funding from the public, who trust that the research conducted is high quality, honest and efficient. Hence, for the future funding and sustainability of research, it is vital that we maintain and nurture this trust.
Image issues in the literature
In May 2020, a Nature news feature introduced readers to the image duplication ‘super-spotter’, microbiologist and scientific integrity expert Dr Elisabeth Bik. Dr Bik, a passionate advocate for upholding research integrity, uses her ‘image forensics’ expertise to identify and report cases of questionable data and image duplications to journals.
In 2016, Dr Bik co-authored a publication that investigated the scale of image duplication using 20,621 biomedical research papers and 40 journals between 1995 and 2014. Alarmingly, 4% (around 1 in 25!) of papers had inappropriate duplications, mainly found in western blotting micrographs and microscopic images.
These startling statistics raise more questions – what about less obvious manipulations of graphical data or images? How many of these examples of inappropriate images are due to research misconduct versus innocent error? These questions remain unanswered and are almost impossible to measure, but they are important to address if we want to ensure future studies are valid, form new research questions, and maintain a positive public perception of science.
Research integrity and COVID-19
Two studies that made headlines during the COVID-19 pandemic reported clinical data investigating the efficacy of hydroxychloroquine and blood pressure medication for COVID-19 treatment. These studies were published in The Lancet and the New England Journal of Medicine, respectively. Both studies used the same outsourced medical database and ultimately, the findings of the articles could not be verified and both articles were retracted at the request of the authors. This situation was troubling particularly because hydroxychloroquine was being promoted as a promising COVID-19 therapy by prominent figures.
Whilst it was fortunate that the authors requested that the articles were retracted, it could have been detrimental to the public’s perception and trust in the scientific literature, especially during a pandemic crisis. In June 2020, the New York Times reported the situation, saying 'the retractions also raise troubling questions about the state of scientific research as the pandemic spreads’. It is impossible to ignore that incredible leaps have been made in medical research during the COVID-19 pandemic, however it is critical that we maintain public trust and confidence in science.
Why do some scientists slip up?
A research career can be a ‘juggling act’- performing research tasks, data analysis, writing manuscripts, applying for grants, training, attending meetings and conferences, outreach activities, and teaching. It can be difficult to consistently manage time and fulfil all of these roles to a high standard. These time pressures can result in shortcuts, less rigorous experimental procedures, and data mismanagement, increasing the risk of unintentional errors.
The ‘publish or perish’ mentality begins early in a scientific research career due to the acute awareness that career progression is inextricable from publication metrics. It is a painful realisation that after years of time, energy, and sacrifices, you may face termination or have limited future opportunities unless you meet this promotional criterion. But is this enough to make an individual compromise their personal and professional core values? One survey attempted to understand this relationship and reported that amongst 315 medical scientists, there was a strong correlation between publication pressure and scientific misconduct. 15% of the students surveyed reported that they had participated in scientific misconduct. A more recent study of 423 researchers (from different disciplines) showed a small correlation where 3% of participants admitted previous misconduct. More worryingly, 51% reported witnessing misconduct from colleagues.
The reasons behind data integrity failings appear to be multi-factorial, although most studies cite professional stress and pressure. In the cases of misconduct, these stressors are often paired with a low probability of getting caught, demonstrating the need for increased research governance.
What can we do about it?
Rejecting hypotheses is just as important to science as accepting them, however, this is not always reflected in our laboratories, institutions, or journals. Publication of positive results in high impact journals is a major performance indicator for both a scientist (by their institution) and their institution (by external bodies). Therefore, it is professionally inefficient to commit time and energy to complete and submit negative (but valid) data for publication. Scientists may also be concerned with the wastage of funding and resources, as these could negatively influence the success of future funding applications. However, it is critically important to understand that this current paradigm is detrimental to not only research integrity, but also reproducibility, funding and future studies.
It is important to highlight that research integrity excellence begins with individuals. The education and guidance of early career researchers on the importance of scientific integrity and the associated repercussions may aid better quality scientific research as they progress throughout their careers. We should place importance on raw data management, accurate laboratory notes for traceability purposes, and creating a secure environment where experimental procedures and group discussions are more transparent.
Journals can help to uphold scientific rigour by giving authors guidance and requiring that they meet certain requirements for experimental design and statistical analyses, for their paper to be considered for publication. The British Journal of Pharmacology (BJP) has done a great deal of work in this area, as outlined in this declaration published in 2018.
We spoke to the BJP’s Editor-in-Chief, Amrita Ahluwalia about this work:
As the world’s leading basic pharmacology journal, the BJP has a responsibility to set standards for transparency and scientific rigour for preclinical research. To this end, we have published a series of guidelines to encourage best practice for natural product pharmacology, animal experimentation, consideration of sex as an experimental variable, design and analysis, data sharing, and preprint citation. These guidelines support authors to use the peer-reviewed evidence base to create robust and reproducible datasets and thus support reviewers and editors to uphold our high standards for publication.
We also actively support authors in regions where we have high numbers of submissions but relatively few acceptances. We have introduced several initiatives including hosting regional webinars to help guide prospective authors through our publication requirements.
Looking forward
No matter how far we progress in our scientific careers, errors and oversights can happen. Having realised that two published academic articles and years of work on a G-protein coupled receptor were derived from an error, Nicola Smith described her academic paper retraction process as “the worst time of my career”. She worried her career and future in academia was over. However, she is now grateful that she was transparent and has managed to move on with her research, albeit with more strict guidelines on keeping laboratory records. Similarly, having discovered that previously published work was not reproducible in the laboratory, Nobel prize winner, Dr Frances Arnold, wrote this tweet. The paper was retracted, and Dr Arnold immediately accepted responsibility for the oversight. The honesty and transparency of both scientists shows commitment to data integrity and should be commended by the scientific community. Hopefully these examples will help to reduce the stigma behind paper retractions.
It is inevitable that errors and paper retractions will occur in the future of research however, it is important that we learn from mistakes and take precautionary measures. Particularly in a global crisis like COVID-19, it is crucial the non-scientific community can trust the work published in scientific journals. Already, the Lancet has adopted more rigorous procedures when handling information from external databases. Software has been developed to detect inappropriate image manipulations and Elisabeth Bik is continuing to battle research integrity issues that exist in our publications. Scientists should adopt transparency measures and leave the fear of retractions behind as we strive to improve the quality of our published work.
Finally, the efficiency and integrity of scientific research could be further improved by promoting the publication of negative data. This would reduce publication pressure, stop unnecessary investigations, and prevent wastage of resources, time, animals and funding.
Comments
If you are a British Pharmacological Society member, please
sign in to post comments.