Research Integrity During the COVID-19 Pandemic: Perspectives of Researchers at an Academic Health Science Center

Abstract

During the coronavirus disease 2019 pandemic, a complex mix of political pressure, social urgency, public panic, and scientific curiosity has significantly impacted the context of research and development. The goal of this study is to understand if and how researchers are shifting their practices and adjusting norms and beliefs regarding research ethics and integrity. We have conducted 31 interviews with Health Science Researchers at the University of Texas Medical Branch which were then analyzed using integrated deductive and inductive coding. We categorized participant views into four main areas: 1) limitations to the research design, 2) publication, 3) duplication of studies, and 4) research pipeline. Although certain researchers were in keeping to the status quo, more were willing to modify norms to address social need and urgency. Notably, they were more likely to opt for systemic change rather than modifications within their own research practices.

Introduction

To mount an effective emergency response to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the traditional methods of research and development have undergone substantial modification since January 2020. A significant infusion of funding has sponsored a network of concurrent coronavirus disease 2019 (COVID-19) research projects on a global scale (Norton et al. 2020; Prudêncio and Costa 2020). Initiatives have been introduced to accelerate work on treatments and vaccines by overlapping various stages of clinical trials, as well as using new screening techniques such as machine learning to identify the most promising antivirals (Gates 2020). In various countries, “challenge trials” are being considered in which participants are intentionally infected to shorten the vaccine test phase and abbreviate the research process (Eyal, Lipsitch, and Smith 2020). Political initiatives like Operation Warp Speed in the United States have created expectations of a quicker research process that may reduce mortality and morbidity rates. While the atypical nature of the COVID-19 research approach does carry with it a higher degree of uncertainty and risk, COVID-19 is a novel virus that may well warrant atypical research methods in the name of the public health emergency.

Knowledge transfer and translation has also undergone a major transformation in the battle against COVID-19. International collaborations have increased and become more interdisciplinary (J. J. Lee and Haupt 2021). Medical journals have increased the speed of peer review for COVID-19 (Horbach 2020). Certain commercial publishers lifted paywalls for COVID-19 research to ensure accessibility of research (Besançon et al. 2021). Increasingly, researchers have been publishing their work in a pre-print format for rapid distribution (Johansson and Saderi 2020). There is also significant public media use of preliminary pre-print trials, even if these studies do not translate into favorable end-phase human studies. While transparency is laudable, premature and potentially misleading announcements may create unrealistic expectations, confusion, and disappointment (London and Kimmelman 2020). The perceived failure to deliver on expectations (e.g. effective treatments) may erode public trust in science. Even when vaccines were deemed effective, the lack of trust in science and medicine resulted in increased vaccine hesitancy (Hornsey, Lobera, and Díaz-Catalán 2020; Goldenberg 2021). Sound health decisions are justified and based on scientific evidence; however, without the acknowledged authority of, and trust in science, it may be difficult to achieve public ‘buy-in’ of recommended health practices and ensure collective order (Singh and Ravinetto 2020, 19; Schwartz 2020).

Since COVID-19 has become the dominant focus of public health research, non-COVID-19 research has been impacted (Sohrabi et al. 2021). Shelter-in-place and physical distancing protocols impede or even prohibit conducting non-COVID-19 studies in health centers or laboratories. Contingency plans had to be made in record time frames; research was modified substantially to reduce travel and proximity of researchers and study participants. When possible, studies have decreased the volume of in-person study visits. Most institutions have had to identify and prioritize the continuation of critical non-COVID research that, if subsided, might cause major direct harm to research subjects, as is the case for research on life-threatening conditions.

According to London and Kimmelman(2020), there should be no “research exceptionalism”; in other words, critical situations like the Covid-19 pandemic should not justify exceptions to the rigors of science. However, while rigor, reproduction, and replication are important in reducing uncertainty in science, they typically lengthen the scientific process. Stakeholders may grow impatient in the face of increasing mortality and morbidity rates. Undoubtedly, researchers are creating innovative mechanisms that will be ethically sound during the pandemic; however, in the rush to achieve or publish results, one should also expect the increased likelihood of unforeseen or unintended problems. For example, The Lancet and the New England Journal of Medicine published studies on Hydroxychloroquine for COVID-19 (Mehra, Ruschitzka, and Patel 2020; Mehra et al. 2020) which when revealed to be unreliable and fraudulent, had to be retracted (Smith 2021).

The goal of our study is to understand shifts in research integrity during the COVID-19 pandemic as observed by researchers at an Academic Health Science Center specialized in infectious diseases. This study aims to identify if and how researchers are (1) shifting certain practices and (2) adjusting norms and beliefs within an institutional setting. It is not intended to develop generalizable knowledge regarding scientific practices, but rather aims to explore how science and scientists are influenced by modifications to their research environment due to the pandemic.

This study was conducted at the University of Texas Medical Branch (UTMB) as it houses a broad array of researchers who conduct both COVID-19 and non-COVID-19 related research. UTMB includes many researchers specialized in infectious diseases who may also access the Biosafety Level 3/4 conditions of the Galveston National Laboratory (GNL) when conducting studies. The GNL is one of only four laboratories in the Unites States that include Biosafety Level 4 (BSL-4) suites designed to engage in rare and valuable research and innovation related to highly infectious pathogens (National Institute of Allergy and Infectious Diseases 2018).

Methods

A semi-structured interview guide was developed based on ongoing literature on COVID-19 research integrity and the knowledge of authors and collaborators, who are applied research ethicists. Interview questions centered on COVID-19 issues specific to research integrity, as well as relational issues that are published in an upcoming paper¹. The full interview guide is available in Supplementary Materials.

Using purposive sampling, we identified the demographic characteristics (gender, age, race) of diverse UTMB faculty who conduct research in different fields including: bench sciences, clinical research, public health, and interdisciplinary or translational science. Three interviewers completed 31 semi-structured interviews that were recorded, transcribed, and anonymized. Interviewers reviewed the transcripts to ensure that they reflected what was said in the recording.

Qualitative coding was conducted in Atlast.ti version 9.1.0. An integrated deductive and inductive scheme was developed by two authors of this paper (CR and ES). This method allowed us to begin our inquiry with specific codes from the questionnaire that were then refined based on the data to ensure that they reflected the thematic concepts suggested by the interviewees. We used Krippendorff’s c-α-binary to assess intercoder reliability for three interviews to evaluate the validity of the coding framework (results in Supplementary Material). Coders agree that quotes chosen for this paper represent the diversity of interviewees’ views.

Results

This study reflects the demographics of the research environment at UTMB regarding sex and race. Our sample had many junior and senior researchers with leadership roles during COVID-19. The definition of “leadership” broadly included chairs of departments, institute directors, educational program directors; faculty had different types of responsibilities in research, education and clinical service. The tables 1 and 2 below outline the demographics of this study².

Table 1:

Demographics (n=31)

Sex
Male	20
Female	11
Race/Ethnicity
Asian	1
Black or African American	3
Hispanic or LatinX	2
Native Hawaiian or Other Pacific Islander	2
White or Caucasian	21
Not disclosed	2
Leadership Positions
No	9
Yes	22
Age
30–40	8
41–50	5
51–60	4
>60	9
n/a	5

Table 2:

Types of research

Field of Study
Basic Sciences	7
Clinical Research	5
Public Health and Health Professions	11
Interdisciplinary and/or Translational Medicine	8
Human Subject Research
No	11
Yes	20
Galveston National Lab BSL 3–4
No	27
Yes	4
Animal Research
No	21
Yes	10
Conducts research related to COVID-19
No	11
Yes	20

We analyzed four topics (codes) linked to research integrity: limitation in research design, publication, duplication, and research pipeline. Detailed definitions of these codes are available in Table 3. For each code, we explored the reasoning for and against shifting norms in research expressed by interviewees.

Table 3:

Description of codes related to research integrity

Code Name	Description	Total Number of Times Coded in Transcripts	Number of Researchers Discussing this Issue
Limitations in research design	Modifications and allowances in research design due to limitations that exist because of Covid-19 or did not exist prior to Covid-19. Limitations may include smaller sample size, lack of randomization, use of unvalidated end points, or other similar limitations made in research design.	52	31
Publication	Discussion regarding the publication process and its impact on science, including: rushed publication, retraction, poor quality of publication, accessibility of published work and availability of research.	66	31
Duplication	Researchers both inside and outside institution simultaneously completing similar or identical research studies, either with or without knowledge that others are also doing so.	28	24
Research Pipeline	The standard process in which research is conducted from start to finish as required by the FDA for approval. Example: biomedical -> pre-clinical -> clinical phase I, II, III.	34	27

Code 1. Limitations of Research Design

We asked interviewees if the COVID-19 pandemic justifies certain modifications or limitations to study design, such as accepting a smaller sample size, lack of randomization, or unverified endpoints. Approximately half of the interviewees explained that the COVID-19 context should not be used as a reason to modify the status quo for their own research. Some interviewees explained that increased limitations are synonymous with “cutting corners.” A few respondents suggested that they believed that research, even before COVID-19, lacked the standards necessary for rigorous research. Certain interviewees referred to rules of rigor as being necessary to get to the truth or understand reality. Modifications to standards would decrease validity and result in untrustworthy research which would simply not benefit science and society and could pose a significant safety concern for human subjects and the broader population. One respondent mentioned that with the increase in COVID-19 funding, researchers should design projects with additional research integrity safeguards to increase trustworthiness.

Some researchers initially stated that they would not change standards but eventually settled on accepting certain modifications, as long as they were very transparent in the study publication. Those who agreed that certain standards could change often specified where changes were warranted and others that were not.

Sample size I’d be willing to sacrifice. Diversity, probably not. Especially since we have all the work being done around racial inequity and the immigration climate in Texas, I would not feel ethically good about an all-white sample for my work. (ID2)

Many researchers mentioned that changing the scale of the study to create smaller exploratory or pilot studies could be justifiable, especially at the beginning of the pandemic when little scientific research regarding COVID-19 had been completed. Interviewees noted that smaller studies are simply less costly and more manageable, and can lead to more rigorous, better justified, larger projects such as randomized clinical trials. Some perceived this as altering expectations of a study that may not have been acceptable in pre-pandemic times.

I have noticed a bit of a relaxation away from the need for a controlled trial amongst some of the more, what we call bench researchers at UTMB where experimental conditions may not be ethical or feasible. We only have funding for 20 tests. It’s a pilot project. We’ll treat it as such and it’s preliminary data, but it’s still research. I am seeing some who I would have thought traditionally before this happened would have totally dismissed such an idea and they’re open to it and wanting to participate, seeing value in it as something that could lead to more research or help provide insights into where to go next. (ID17)

Some interviewees suggested that the level of rigor or standard of evidence applied in the clinical space is not narrowly based on one study but rather on the totality of studies regarding a topic. As such, certain researchers believed that the shortcomings of one or few research studies would likely get invalidated before being translated to any practical user-based application.

Many researchers used the example of vaccine development to show that while standards may have been modified and research fast-tracked, sufficient safety standards have been maintained. Certain researchers mentioned that the cost or risk of not having a vaccine was more problematic than fast-tracking research via expanded use access (EUA). One interviewee suggested that we should be looking at diversifying the research portfolio as opposed to overemphasizing increasing the pace of vaccine studies. Specifically, the interviewee suggested that effective polymerase chain reaction test (PCR) development and infrastructure testing should have been the subject of more expedited scientific studies.

People in our lab and labs like this around the world, work with sequences all the time. They can design and test PCR technology very quickly. So rather than having this centralized at CDC³ and FDA⁴ squelching any effort by local laboratories to provide diagnostic testing in it, at a time when there were a few cases and contact tracing could have been done and we could have had a much better control effort in place. I think that was a very serious mistake, but that’s... So that’s where I would have cut corners. (ID15)

Although many researchers focused on COVID-19 research, those researchers doing non-COVID projects suggested that they faced important limitations and had to adjust their work. Non-COVID-19 related studies were often funded to be completed within a specific timeframe. With the shutdown, social distancing, and limited human subject recruitment, researchers could not necessarily complete their study according to original methodology and protocols.

Another respondent mentioned that all studies conducted during COVID-19 should include COVID-19 considerations in the research budget (additional replication, data, triangulation resources). If COVID-19 or a history of COVID-19 is part of all studies, – another determinant of health of sorts - like gender, sex or race – this will help in understanding COVID-19 impacts on other diseases. For example, in a study evaluating new asthma treatments, one should also include the COVID-19 test results as a data point. Even if the main objective is not related to COVID-19, acknowledging, and understanding its collateral and downstream effect on various systems would be valuable. Similarly, one respondent mentioned that studies that were developed pre-COVID to discover a non-COVID phenomenon may need to be limited to COVID-19 negative populations until we truly understand the impact of COVID-19 on an individual’s body. Interviewees did mention that this could make recruitment more difficult and samples smaller (for overview of Code 1 see Table 4).

Table 4:

Overview of justifications for modifying, or not, research design during pandemic^*

Reasons safeguarding norms	Reasons for modifying norms
• Norms are already not rigorous enough • Rigour, validity, reproducibility • Harm to participants or population • No downstream benefit to population • Loss of sense of reality • More funding allows for more integrity safeguards	• Risk of not having knowledge is more dangerous • Certain norms are not essential in some studies • Increasing pilot/experimental studies to establish preliminary knowledge. Such studies do not always respect the norms of science • Speed

^*this table presents norms in no particular order of importance

Code 2. Publication

According to most interviewees, COVID-19 accentuated the rush to publish with the goal of establishing a knowledge base that could help in developing public health policies, clinical best practices, and basic biochemical understanding of COVID-19. A few researchers did mention that the increased speed of production and publication does not impact the integrity of the science. These researchers explained that they increased work hours while also cutting down on their personal activities sacrificing work-life balance during the pandemic.

Many researchers recognized that the high potential for innovative COVID-19 research is an opportunity for success in an already competitive environment. According to interviewees, this competitive environment seemed to be driven by many stakeholders, including the general public’s desire for information, the research institution managing career advancement, and even the journals publishing scientific. Many interviewees explained how this competition, opportunity and pressure can be at odds with quality and rigorous research. Most did not believe that there was intentional, pre-meditated wrongdoing. Some mentioned that research has become slightly “sloppy” or that researchers were not as careful as they had been in the past. A few mentioned that researchers are intentionally “publishing garbage” (ID 24). For some, this low quality is a passing phase due to psychological pressures of the pandemic:

I’ve seen papers that in my opinion were premature. And I think it was just because of the pressure. I mean, again, I think that these phenomena are transient my opinion within maybe six to seven months, once the vaccine is distributed everywhere pretty much in the US, then I think that these phenomena or maybe not super rigorous science is gonna fade away. And there is also one thing that is very important, it’s we are all humans, okay? So there are a lot of psychological factors that make us do things. So even as researcher we might have fears, we might have lost someone, we really might have been in... maybe we do things because we really feel it’s not really that we are dishonest, but maybe we really feel the pressure of just saying something. (ID10)

Some mentioned that it is not the data or analysis but rather the conclusions or generalizations in manuscripts that are very problematic. This was exemplified by one interviewee who explained that public health journals sometimes require policy recommendations. However, studies were often so small that any generalized recommendation seemed premature and important limitations were ignored. Such generalizations could lead policymakers to base their decisions on preliminary knowledge that is not and should not be deemed conclusive.

Many researchers discussed the uptake in non-peer-reviewed publications or datasets; this includes repositories as well as pre-print archiving (e.g. bioRxiv, medRxiv). Several mentioned using repositories to find information at the onset of the pandemic since it was the only thing available. Conversely, many also expressed serious reservations with repositories: notably, the lack of due process, quality control, as well as problems with interpreting the results of pre-prints. Although researchers are trained to consider certain data and research results with a degree of healthy skepticism, interviewees did worry that the untrained reader may have more trouble discerning the limitations of preliminary research. Certain interviewees opined that research should not be considered “science” before it is properly peer-reviewed.

A few interviewees who also had editorial responsibilities identified the lack of competent peer-reviewers for COVID-19 research as an important limitation to quality publication. Since peer-reviewers are generally researchers with significant knowledge in a field, they may also be busy with their own COVID-19 research and have limited time to peer review. The same interviewees also revealed the motivation of editors and publishers to be the first to present a new finding for prestige and financial gain. A few other researchers emphasized how the prioritization of COVID-19 research in journals may have gone so far as to essentially control the scientific message.

I did a meta-analysis with students on Statins. Our timing was such that it was submitted to JAMA, I don’t know if it was March or April. And it was rejected, because JAMA is not accepting any papers that are not about COVID. I mean, that’s a dramatic change for JAMA. So that is a COVID-only journal right now. And it’s interesting, because we still have that substrate of chronic diseases going on in society, but I think we all know we’re looking at it, I mean, the impact on elderly, on households. It’s because of COVID and the economy because that’s the other thing that’s happening to everybody. (ID28)

With the prioritization of COVID related publications, non-COVID interviewees felt like they were left to compete for less funding and fewer publication opportunities. In reaction to current publication controls, new alternative journal models have emerged that promote a more constructive feedback process.

I think that there are some new journals like eLife which is a journal that is really run by scientists. It’s interesting because the review process involves the reviewers, after they send their written reviews, you’re supposed to get on a teleconference with the other reviewers. First of all, it makes it more civilized than some of these anonymous review comments were so nasty that you wouldn’t do that if you know you had to sit in a televideo and defend yourself, even though it’s not the authors but it’s another colleague in the field, so I think it’s become a more civilized process. (ID 24)

Interviewees recognized the need to be hypervigilant and critical of problematic publications. Some researchers mentioned that they increasingly can’t trust the findings or conclusions of published works. Although some interviewees were a little more flexible about quality norms or limitations, public transparency respecting research evidence, analysis and design remained of central importance (overview of Code 2 see Table 5).

Table 5:

Mapping Publication Issues and Benefits during COVID-19 pandemic^*

Publication Issues during COVID-19

Publication Benefits during COVID-19

• Reduction in quality and rigor • Premature publications • Uptake in non peer-reviewed resources (publication, databases) of questionable quality • Uncertainty regarding non peer-reviewed work • Editor and publisher conflict of interest • Journals are less likely to publish important non-COVID work • Duplication leading to less efficiency and novelty • Chaotic knowledge translation because of duplication

• Increase in certain journal standards • Increased opportunities for novelty regarding COVID-19 research • Quicker publication process (quicker peer-review) for COVID-19 research • Increased discussion to promote transparency regarding limitations • Duplication of similar research leads to more replication increasing long term quality

^*This is a summary of issues and benefits in no particular order of importance. This includes publication issues linked to code 2 (publication) and code 3 (duplication).

Code 3. Duplication

We asked interviewees if the increasing number of COVID-19 research publications would cause duplication and if this was problematic. While some expressed reservations about duplication, many researchers found it justified to ensure that science is reproducible. Certain researchers specifically mentioned that duplication could allow for a more representative viewpoint in clinical research if human subject samples remained slightly different from one study to the next. Others explained that simply having different research teams conducting the research will yield different interpretations, methods, and results.

Researchers often acknowledged the significant need for research on COVID-19 and emerging infectious diseases that may lead to epidemics or pandemics. However, it was very difficult for researchers to determine how much research on a topic is sufficient. At some point, it becomes a question of resource availability and prioritization. Many interviewees did mention that duplicating research had become a waste of research funds but that certain bad decisions are to be expected during the chaos of a public health emergency.

Certain researchers gave specific reasons why they find wasting resources to be truly problematic. One researcher considered repeat studies on animals to be a waste of animal lives.

I think if it’s a really clear duplication, and we do a lot of animal work, I see in animal welfare an issue with this. We are always trained to keep the animal numbers down as much as possible. If you repeat studies, you’re wasting animal lives. (…) I think we’ve all seen the studies where it says 80% of all the experiments cannot be repeated or so. (…) One could argue and say, “Well, I repeated this in my lab and yes, it works or no, it works slightly different.” I think one could argue that way, but I don’t know. Especially when it comes to animal lives, I’m not a big fan. (ID 6)

Another researcher also described the issue of duplication in human subject research. Instead of finding practical interventions that may reduce or resolve health issues, we seem to duplicate the same types of observational studies. Specifically, the interviewee explained that the same research with human subjects is often duplicated to demonstrate or confirm health disparities affecting mainly vulnerable populations without further exploring practical approaches to address these disparities.

… I am concerned that it will be obvious what the impact of COVID is on vulnerable populations, yet we will continue to study it without funding the existing evidence-based solutions that we could implement now. (ID 2)

One interviewee mentioned that the disorganized duplication of COVID-19 research may have confused the public about scientific advancements. A few researchers highlighted ways to avoid duplication including better organization, increased collaboration and more effective institutional or systemic use of resources (mainly human biospecimens).

Code 4. Research Pipeline

We asked if the pandemic justified changing the typical process of research and development. This process is often viewed as linear, extending from pre-clinical, to animal studies, to clinical phases 1, 2, 3, 4, and finally implementation. Although four researchers mentioned that the scientific development process should never be modified, the majority of researchers (N=23) suggested that modifications could be warranted. Regardless of their views, most identified the negative impacts and outcomes that may result from altering the system. Safety concerns for both human subjects and the general population were one of the main concerns.

We have a process for a reason. (…) I know this is a matter of life and death. But if we do it right, we are less likely to kill people at the end. Then we could kill them with the vaccines, because we combine things [steps of clinical trials]. So, no. I believe in sticking to the protocol, sticking to what we know to be best practices. Maybe someday we’ll find out what we did was wrong. But right now, based on our life experiences as researchers, this is best practices and we need to stick to it. (ID 7)

During the interviews (September-December 2020), a few interviewees mentioned that the vaccine data for Pfizer and Moderna do not include long-term testing data to fully explore and effectively mitigate long-term problems. One researcher admitted that the idea of fast-tracking research was justifiable from a scientific standpoint, but that it still made them very uneasy, admitting to having vaccine hesitancy and preferring to wait before getting vaccinated. Efficacy seemed important to many but certainly secondary to safety.

Certain interviewees mentioned that it would be challenging to significantly modify the process because the ensuing questions and uncertainty might further reduce public trust in science. One researcher suggested that there is an obligation to be even more rigorous at this time when public trust in science is precarious.

Certain researchers feared that adjusting the research process to meet COVID-19 expectations would generally heighten future expectations. Many researchers did not want to apply the streamlined COVID-19 research process to research of non-pandemic illnesses.

I think, now that I think maybe the general public in particular, has seen what happens when all resources are on one topic and all of the brain energy is on one topic, like what is actually possible in terms of how quickly things can get to get done, that will certainly, I think change the level of expectation the next time there’s a public health crisis. Well, science will be able to come up with a solution within nine months to a year. So, I don’t think that dramatic changes to regulation and oversight that have happened because of COVID would need to be broadly applied to all areas of research. I think it could be done on a kind of a case by case basis. (ID 4)

Even if many researchers didn’t wish to radically modify the process, many did identify bureaucratic steps that are not scientifically necessary and could be altered or removed altogether. They mentioned that there is a lot of “fluff”(ID 5) or “red tape” (ID14). Certain researchers discussed how the production of novel vaccines may appear to have been streamlined due to Operation Warp Speed, but in fact, messenger ribonucleic acid (MRNA) vaccine technology, developed long before the pandemic, was a major factor in increasing the speed and pace of the process.

I think a little different because actually just COVID-19 the mRNA technology and ideas were already there. So maybe they didn’t have to go directly back to step one, like they did back in the nineties when they created this. But I do think that potentially for an emergency where there’s a need and 3000 people die in a day in this country, as an example, potentially combining phase one through phase three, we talked about that sort of with the FDA review at least a little bit. I think it’s reasonable to combine the phase one, two and three and just do one very quick 30,000 trial. And then say from there we have some data, even if limited, but that’s enough to probably say it’s safe enough to give a try. So I think yeah, if there was an emergency virus like this, which is killing so many people a day across the world, it’s reasonable to maybe on the later end, not do as many reviews. (ID14)

This previous interviewee, like many other interviewees, justified speeding up the process based on the ever-increasing mortality rate. Some interviewees were more wary about expediting vaccine development, because they were concerned about lack of accountability for issues that may arise. A few researchers went deeper to explain how procedural accountability could be ensured using continuous quality control. One researcher explained that we can analyze the research process itself to see what worked and if it was justified. However, the planning and resources (financial, scientific, human) to reconsolidate the process would be substantial and therefore even if we could change the system, interviewees were not sure we should, especially beyond the COVID-19 emergency.

I think you can consolidate that efficiently and do speed up the process without impacting the quality. (…) It’s going to be more expensive and a lot of people working really hard to make it happen, but I think we’ve seen that it can be done and probably should be done. Should it be done every time? I don’t know. It depends on the product and the urgency. (ID15)

One researcher mentioned that we should be changing the traditional ‘pipeline’ process. However, they said it should not be done simply to increase the speed of process, but rather to extend the research to include and apply to a greater diversity of the population.

Has COVID changed the world? I think it’s made it even more imperative that we’re safe. Particularly for a vaccine, for a disease that’s deadly, already deadly in some parts of our population. And the disparity with which people have been suffering from COVID-19 adds to that. So, it almost seems to me that in the standard pipeline, we didn’t pay much attention to sex as a variable, or race as a variable, in terms of what made it through that pipeline. And how are we going to deal with that? Or chronic comorbidities, how are we dealing that with the vaccines? (…) there are some things you can skip, and there are some things you can’t skip. (ID 18)

Another researcher explained that greater speed in the pipeline may be done by repurposing already accepted medications and treatments – including those with monoclonal antibodies. Researchers did explain how such treatments are relatively safe in terms of toxicity but require substantial effectiveness testing (overview of Code 4 see Table 6).

Table 6:

Arguments for and against systemic change^*

Arguments against system change	Arguments for system change
• Fear of creating false expectations for the public • Fear of needing to apply new standards to all studies • Safety and efficacy concerns • Belief in the process – uncertainty in new system • Resource intensive	• Speed • Exceptional situation • Increase diversity (translation of system)

^*This table maps out arguments in no particular order of importance

Discussion

The scientific literature has been quick to identify the shortcomings of modifications of the status quo during the pandemic, often pointing out issues of safety and efficacy (London and Kimmelman 2020). Many journals put out editorials and commentaries underscoring the importance of ethics and integrity regarding COVID-19 studies and non-COVID studies to protect wellbeing of participants and mitigate risks to study integrity (Fleming, Labriola, and Wittes 2020; Dinis-Oliveira 2020; Bramstedt 2020; Bagdasarian, Cross, and Fisher 2020). Such papers, while important, do not consider the real-world complications of the various research environments.

Our paper contributes to the literature by exploring the practical concerns, suggestions, and reasonings of frontline researchers at our institution. Some maintained the status quo found in the literature pre-pandemic; and when they did, they often argued based on fear of not upholding well-known principles of science. In our study, there was a minority viewpoint that research is equivalent to “the truth” and as such, any change would erode some veracity and diminish reality; this is more akin to a belief system or faith. Although it is important to promote rigorous practice, understanding that scientific results are never an exact mirror of reality seems to have been ignored in the overly positivistic views of certain researchers. Various philosophers and social scientists have demonstrated that science does change based on the method, vision, and social and power dynamics (Kuhn 1962; Latour 1996). Although most of our interviewees admit that science needs to be quick and responsive because of the pandemic. Knowing if and how scientific methods, processes, and the system itself should change remains an important yet challenging question.

Interviewees who argued for the status quo created a good prima face perspective built on caution and clinical best practice. Changing this prima face status quo would require practical and logical justification. Interestingly, when interviewees did want to change certain aspects of research, their explanation for doing so was often very detailed and rationalized. This may be due to interviewers probing new interview discussions; however, it may also be that researchers feel an obligation to justify and maintain their work as rigorous and scientific in an atypical setting.

The types of change deemed acceptable by researchers were often shaped either by their local research context, or from a more systemic perspective. For example, certain researchers would never limit diversity of their sample given the importance of social discrimination in their research context. Such a specific research-based change is locally driven and may be specific to the type of research, the field, and the object of research. Other interviewees justified changing the status quo through systemic modifications, such as overlapping certain phases of clinical studies because the typical process is slow. Although this study is by no means generalizable, there was a tendency for researchers to opt for systemic change (in code 4 on research pipeline) in comparison to code 1 (limitation to the research design). See Figure 1 for an overview of how our study codes span context and broader systemic considerations.

Figure 1 : Overview of codes throughout the R&D process.

Although there was little consensus among interviewees in our survey sample as to how to speed up research safely, there were topics that seemed to attract greater interest. One recurring discussion centered on a desire to increase the number of “pilot” or “exploratory” studies in which researchers could transparently use methods that were quicker, different, innovative, or ground-breaking. Publication and dissemination of exploratory study information does raise many questions; should exploratory work be attributed the same scientific status as a peer-reviewed study? If not, is it acceptable to share “half baked” research if it could help public health? Although some mentioned that public health decisions are often made based on limited evidence, researchers are hesitant to allow communication of findings or results that are not up to scientific standards.

Even before the pandemic, competition in the system of science has been shown to incentivize priority of discovery resulting in smaller sample size, reduced reliability and less information per publication (Tiokhin, Yan, and Morgan 2021). To reduce the amount of smaller (pilot or exploratory) studies, interviewees highlighted the importance of collaboration to effectively pool resources.

Collaboration and reorganization of scientific resources has been promoted since the beginning of the pandemic. For example, the US National Institutes of Health has coordinated public and private researchers globally to reduce duplication and focus on promising vaccine and treatment resulting in the Accelerating COVID-19 Therapeutic Interventions and Vaccines (ACTIV) partnership (Sampat and Shadlen 2021). However, beyond NIH partnerships, important challenges to collaborations have reduced the average team size. Studies of team creation from the literature published during autumn 2020 demonstrated that COVID-19 research teams are smaller (measured by number of authors, nations and rate of international-teamed articles) when compared to pre-COVID-19 times (Cai, Fry, and Wagner 2021; Evans et al. 2020); this has been explained by travel bans and political tensions mainly between China and the United States (Maher and Van Noorden 2021). Unfortunately, team environments are also less diverse as women and underrepresented minority researchers are disproportionately affected by disruptions linked to the pandemic (Carr et al. 2021).

Moreover, collaborations seem to have been technoscientific in nature and often limited to laboratory sciences and clinical research with little broader interdisciplinary consideration that would include social sciences and the humanities (Ellemers 2021). This was exemplified in our interviews when researchers mentioned that speeding up vaccine development is one thing, but ensuring collective cooperation to ensure widespread vaccination is another issue. There certainly have been a significant amount of social research on topics (e.g. vaccine hesitancy, social isolation impact, social justice); however they do not seem to be developed in a collective interdisciplinary manner to achieve effective preventive public health interventions (Aristovnik, Ravšelj, and Umek 2020).

Another area that garnered concern by interviewees focused on the quality of published work that notably lacked peer-review. This worry is echoed throughout the scientific literature on publication ethics, underscoring its impact on the quality of papers and number of retractions (El-Menyar et al. 2021; Abritis, Marcus, and Oransky 2020). Although peer-review is considered a quality control mechanism, most empirical studies of the peer-review process itself, even before the pandemic, identified a significant bias inherent in peer-review (Smith 2021; C. J. Lee et al. 2013; Resnik and Smith 2020). One element that peer-review and editorial review can provide is structure through editorial policies. A rapid systematic literature review of the COVID-19 publications conducted by Sullivan and colleagues suggested that compliance documentation was lacking at the beginning of the pandemic even with editorial policies in place (O’Sullivan et al. 2021). Another study by Li and colleagues (2021) highlighted the low quality of more than half of the systematic reviews.

However, pointing to peer-review as the ideal way forward to ensure the quality and integrity of published research is problematic; two individual peer reviewers cannot fully assume responsibility or accountability of a research team. In our study it was often suggested by researchers that publication makes a paper “scientific”. This may symbolize the importance of critical feedback but also perpetuates procedural legitimization and conservativism (Godlee 2002). In other words, a researcher seems to need a process that differentiates “publication” from posting a blog or soapbox. In the literature, conservatism has often been linked to peer-review but it is also linked to restrictions in funding and lack of epistemological freedom (Bedessem 2021). Although there is a significant amount of COVID-19 funding, it was certainly linked to specifically applied translational results that may have reduced the amount of novel and risky research.

Interviewees often discussed many forms of “control” of scientific work when discussing prioritization of COVID-19 research throughout the system (eg. in funding, editorial bias, peer-review). This notion of “control” is similar to the literature on conservatism in social epistemology of science. Conservatism is often defined as opposite to novelty or demonstrating that certain forces constrain or limit the types of problems or question addressed by scientific societies(Currie 2019). Paradoxically, interviewees certainly felt that COVID-19 may have novel issues but that the system confined them to work on certain topics and use certain resources in certain ways. By focusing on the increase pace of research into specific topics such as vaccination, and select treatments, researchers felt unable to diversify or expand their research portfolio, thereby limiting their creativity.

The discourse regarding speed or increased pace of research tends to overlook the direction, (de) prioritization of the research subjects or fields more broadly.

Although total research funding during COVID-19 has increased and diversified with governments, nongovernmental organizations and philanthropies investing in research and development, different types of controls limited research diversification and creativity including funding limitations, publication process, prioritization of observational studies instead of more interventional or targeted studies.

Limitations

Given its qualitative nature, this study is not generalizable. The first researchers to respond to email recruitment request were white men. A higher number of recruitment emails were then sent out specifically to women and racial minorities to diversify the sample. Although only 7 (22%) of interviewees are racial minorities and 11(35%) of interviewees are women, this is actually a close statistical representation of the total population of faculty at UTMB (UTMB 2022). Although there is underrepresentation of the Asian population in our sample, the qualitative and exploratory nature of our study does not require exact statistical representation. Further research should be conducted to understand the specific struggles of racial minorities. The overrepresentation of researchers in leadership positions may also neglected the full experience of junior faculty throughout various scientific disciplines.

This type of exploratory research would be impossible to reproduce at another point in time since it represents a “snapshot” of researchers views at the beginning of the the pandemic. Although this limits reproducibility, it also has significant value as it considers the complexity and impact of extreme uncertainty. Our sample being limited to UTMB does not take into consideration the realities of institutes that have no infectious disease research and may have shut down for longer periods of time which significantly hinders scientific productivity in a different manner.

Another bias that was recognized, only after our data collection, was the overemphasis on COVID-19 vaccine examples. When probing participants, interviewers often provided examples from the media. At that time, various phases of COVID-19 trials were being done quicker often in an overlapping manner. This may have influenced interviewees to focus on vaccine trials and deflect from other concerns respecting COVID or non-COVID research.

Conclusion

This study aimed to understand if and how researchers have shifted or are shifting certain practices and norms regarding research ethics and integrity during the COVID-19 pandemic. We analyzed four main topics, namely: research design, publication, duplication and research pipeline. At the beginning of the pandemic, the public media and certain scholars were mentioning that urgency would result in questionable scientific research during the pandemic (Fleming, Labriola, and Wittes 2020; Roy and Mathew 2020; Dinis-Oliveira 2020). Indeed, there was controversial research, such as Hydroxychloroquine and Ivermectin, that caused concern among researchers. However, this may not reflect the majority of scientific practice during the pandemic.

Researchers did note that shifting scientific norms was, unfortunately, not always done for the “good” of science. Many highlighted the intentional or unintentional neglect of research and poor quality of science mainly surrounding publication and retraction. However, most researchers did not perceive “keeping the status quo” to be synonymous to “good science”. Researchers believed that the pandemic warranted evolution and adaptation to respond to the needs of society. This evolution was not perceived by researchers to compromise their work. Often, they would justify modifications based on the type of research that they are doing. Other times they suggested that the system could change in a way that would promote efficiency. It is important to note that researchers were thinking about ways to make science better, not only to promote quicker science but to ensure diversity of sampling and translation of scientific results. Researchers believed that they were adapting to complex circumstances, which is an important characteristic of professionalism, quick-thinking, and survival.

Given the fact that this study took place at the onset of the pandemic, it is reassuring that researchers were so enthusiastic about thinking about such issues. Most did end the interviews wondering what other researchers thought. Some also mentioned that they had not yet given much thought to important issues that were not directly linked to their research.

Given such interest, there is certainly a need to create an open forum where researchers can discuss acceptable and unacceptable shifts in scientific norms during a pandemic. Any modifications will be shaped by emerging pressures and changing priorities; the ethical foundation in support of change will likely be stronger and more readily endorsed through an iterative and collective debate. Researchers would also have an opportunity to think reflexively and examine their own beliefs, judgements, and practices, and understand the influence those have in their decisions and the research they conduct. Finally, dialogue and collective agreement could be an enabling mechanism to implement systemic changes widely recognized as acceptable, professional, and expert-driven.