Fluent Essays

https://link.springer.com/article/10.1007/s11356-017-0619-5 RESEARCH ARTICLE Good and Bad Research Collaborations: Researchers’ Views on Science and Ethics in Global Health Research Michael Parker*, Patricia Kingori The Ethox Centre, Nuffield Department of Population Health, University of Oxford, Old Road Campus, Oxford, United Kingdom * [email protected] Abstract There has been a dramatic rise in the scale and scope of collaborative global health research. A number of structural and scientific factors explain this growth and there has been much discussion of these in the literature. Little, if any, attention has been paid, how- ever, to the factors identified by scientists and other research actors as important to suc- cessful research collaboration. This is surprising given that their decisions are likely to play a key role in the sustainability and effectiveness of global health research initiatives. In this paper, we report on qualitative research with leading scientists involved in major interna- tional research collaborations about their views on good and bad collaborations and the fac- tors that inform their decision-making about joining and participating actively in research networks. We identify and discuss eight factors that researchers see as essential in judging the merits of active participation in global health research collaborations: opportunities for active involvement in cutting-edge, interesting science; effective leadership; competence of potential partners in and commitment to good scientific practice; capacity building; respect for the needs, interests and agendas of partners; opportunities for discussion and disagreement; trust and confidence; and, justice and fairness in collaboration. Our findings suggest that the sustainability and effectiveness of global health research collaborations has an important ethical or moral dimension for the research actors involved. Introduction The idea that researchers should work in collaboration, rather than in isolation, has signifi- cantly reconfigured medical research. Reflectinga wider trend in scientific research since the Second World War, the scale and scope of global health research has grown dramatically in response to calls for more efficient and collaborative knowledge production [1, 2]. From the late 1990s and early 2000s, additional impetus has been given to these developments by claims that global health research collaborations have the potential to play a significant role in addressing global health disparities [3, 4]. The growth in global health research collaborations PLOS ONE | DOI:10.1371/journal.pone.0163579 October 13, 2016 1 / 19 a11111 OPEN ACCESS Citation: Parker M, Kingori P (2016) Good and Bad Research Collaborations: Researchers’ Views on Science and Ethics in Global Health Research. PLoS ONE 11(10): e0163579. doi:10.1371/journal. pone.0163579 Editor: Anny Fortin, Dafra Pharma, BELGIUM Received: August 3, 2015 Accepted: September 12, 2016 Published: October 13, 2016 Copyright: © 2016 Parker, Kingori. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: Illustrative quotations from the qualitative interviews conducted for this study appear in the paper to support the analysis. Research ethics committee approval was not obtained for the sharing of data beyond the publication of illustrative quotations and consent for this was not obtained at the time of interviews. However, the authors are strongly supportive of data sharing and would be happy to discuss the sharing of anonymised transcripts with interested researchers on a case by case basis. Researchers interested in accessing the data are strongly encouraged to contact the authors directly. http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0163579&domain=pdf http://creativecommons.org/licenses/by/4.0/ has also been driven by scientific developments, including those in genomics and in the devel- opment of statistical techniques and technological requirements for the analysis of very large datasets, and by major funding initiatives such as those of the Bill and Melinda Gates Founda- tion, and the Wellcome Trust [5]. The rapid growth in global health research networks involving complex historical, political and economic partnerships between researchers has been accompanied by literature addressing the ethical issues arising in global health research. Much of this literature has tended to focus on practical ethical concerns in operationalising international guidelines related to valid con- sent, standards of care, post-trial access, acceptable levels of risk, benefit sharing, and so on. There is also an emerging literature on the sociologicaland political aspects of such research collaborations primarily focused on understanding how macro level collaborations function [6, 7, 8, 9, 10]. The starting point for these accounts is a view that simply measuring the growth of collaborations provides little insight into their qualities, or whether they are equal or fair [11]. This literature has tended to suggest that research collaborations reproduce relationships and conditions which disproportionately favour high-income countries and institutions [12, 13]. Publication authorship, the named principal investigators and grant holders for funding appli- cations, staff remuneration policies, tax exemption for foreign researchers and the ownership of samples and data have all been presented as areas where current inequalities undermine equal partnerships and collaborations [14, 15, 16]. It has been argued, that the term ‘collabora- tion’ can sometimes mask or exacerbate such problems [17, 18]. For instance, a recent exami- nation of publications, resulting from global health research collaborations between American and European scientists and those in Central Africa, found that the most frequent role for Afri- can scientists was in providing samples and conducting fieldwork not research design [19]. These debates have been accompanied by a growing literature on the role of ‘ethics governance’ in ensuring high ethical standards in global health research collaborations [20, 21, 22]. Such insights have highlighted the importance of distinguishing between well-managed col- laborations and ethically sound collaborations and have prompted calls for changes to the cri- teria for evaluating and assessing the fair and ethical conduct of such collaborations, such as the extent to which the collaboration provides tangible and timely resources to Global South collaborators and opportunities to be involved in areas of science which can provide opportu- nities for developing and strengthening new skillsets [20]. This academic debate has been mir- rored in a wider policy and political debate about what might constitute the responsible conduct of collaborative research. At the level of guidance and policy, a number of initiatives have attempted to characterise ‘good’ collaborative research practice. These include: ‘Responsi- ble Conduct in the Global Research Enterprise’ [21]; ‘Where there is no lawyer: Guidance for fairer contract negotiation in collaborative research partnerships’ [22]; ‘The Montreal State- ment on Research Integrity in Cross-BoundaryResearch Collaborations’ [23]; ‘The COHRED Fairness Index for international collaborative partnerships [24]. What is striking about these academic debates and policy-level initiatives and guidelines, is that very little attention has been paid to capturing the experiences,values and attitudes of research actors directly involved in such collaborations or to how these factors influence per- ceptions of and judgements about whether to participate in them. Furthermore, they rarely involve consideration of researchers’ perspectives or evidence about the criteria they them- selves employ to assess potential collaborations. As a consequence, there are few accounts of researchers acting as agents, in either constructing their own assessments and evaluations of collaborations or, actively choosing who they collaborate with on that basis. This is an impor- tant omission as an understanding of the judgements researchers make about whether to join or to contribute actively to global health research collaborations is going to be a major factor in understanding the requirements for successful, sustainable science. It is vital for research Good and Bad Research Collaborations PLOS ONE | DOI:10.1371/journal.pone.0163579 October 13, 2016 2 / 19 Funding: This work is supported by Wellcome Trust (wellcome.ac.uk) grant number (96527). Patricia Kingori is supported by a Wellcome Trust Society and Ethics Research Fellowship (095889). The funders had no role in the study design, data collection, analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. funders, researchers, and policymakers to have a good understanding of the experiences of research actors involved in global health and how they reach decisions about collaborations. As a first step towards addressing this gap, we report here on our analysis of a series of interviews conducted with research actors and funders involved in several high-profile international global health research networks. Methods The findings reported in this paper are based on 22 qualitative interviews conducted with research actors over a two-year period. Those interviewedwere purposively selected because they play a significant role in at least one—usually several—large international collaborative networks involving partners in both high and low income countries. The term ‘large interna- tional collaborative network’ here refers to collaborations involving partners in many i.e. more than 10—sometimes more than 20—countries across both low and high income settings. Inter- viewees were selected by their role and ranking in research. Using a snowball approach to sam- pling, a concerted effort was made to include a range of people playing roles of key importance in collaborative research. These included: principal investigators, research funders, network coordinators, clinical trial managers, epidemiologists, laboratory and sample managers, IT spe- cialists, database managers, statisticians, clinical researchers and managers of research and health institutions. The number of principal investigators interviewedwas 11. Interviewees self-identifiedas being from South East Asia, South Asia, East Africa, West Africa, Europe, Oceania, and North America. Each quotation in this paper is followed by a note of the role, self-identifiedgeographical region of origin and current primary location on the interviewee.It is noteworthy, however, that current primary location was in practice often difficult to nomi- nate because the roles played were often very mobile. A researcher might, for example, work and spend significant time in and even consider themselves to live in, more than one location. It is also necessary to note that congruence between region of origin and location of employ- ment does not mean people are working in their country of origin: ‘West African Principal Investigator, West Africa’ might, for example, signify a Ghanaian researcher working as a PI in The Gambia. Attention was also paid to the gender of those interviewed;nine of the interview- ees were female, thirteen were male. Gender is not associated with individual quotations in this paper in order to protect the confidentiality of the participants. Following an in-depth discussion about the aims of the study, data management, and the use of anonymised interviewextracts in publications, participants provided oral consent to par- ticipate. As is common in interviewingsenior professionals, oral consent was deemed appro- priate because the intervieweeswere senior and experienced.Consent was recorded before interviews started and is included in transcripts. This study, including the approach to consent, was approved by the Oxford University Tropical Medicine Research Ethics Committee (Ref: 020–06). Face-to-face interviews were conducted by MP in locations convenient to the inter- viewees. The interviews were about an hour in length. All but two were audio recorded and transcribed verbatim. One interview was conducted by email in French via a translator with key themes verified and expanded upon in person at a later date. The other was not recorded because of problems with the recording device, but detailed notes including verbatim quota- tions were taken during the interview. The interviews covered a range of subjects, some of these were initiated by the intervieweesand some were informed by a topic guide which focused on the following areas: personal research career and experience of collaboration; expe- riences of deciding for/against collaboration; views about what makes the difference between effective and ineffective research collaborations; and factors of particular importance in collab- orations between high and low-income countries. Good and Bad Research Collaborations PLOS ONE | DOI:10.1371/journal.pone.0163579 October 13, 2016 3 / 19 Analysis was conducted in tandem with data collection, with initial analyses of early inter- views informing the themes explored in those that followed. Following the completion of the interviews,analysis was subsequently conducted in a number of stages. The interview tran- scripts were read through by MP and PK and broad themes identified by each author. These themes were discussed and formed the basis of more in-depth line-by-line analysis of the tran- scripts. This was conducted and managed in NVIVO (Version 10). New themes identified in more detailed analysis were used to inform the analysis. An example of a broad theme was “Rule of collaborations” with “Obligations”, “Sticking to the rules” and “Breaking the rules” identified as sub-themes. This analysis was also informed by the literature on this subject and in particular the social science, ethics and medicine publications examining collaborations, partnerships and networks. Findings [. . .] I get invited to participate in many things and in some cases I do kind of turn down the invitation or drag my feet somehow. It mainly depends on you know [if]the science is some- thing [worth doing]. I need to be convinced that it’s worthwhile. (West African Principal Investigator, West Africa) As the number, scale and complexity of research collaborations have increased, individual researchers, such as the informant quoted above, have experiencedmore frequent invitations to join them. For busy researchers—short of time, energy and other resources—this means that decisions, which are sometimes difficult,need to be made quickly about which invitations to accept. Box 1 below sets out a summary list of key factors identified in our research as influenc- ing researchers’ decisions about initiating, joining, remaining in, and actively contributing to research networks and their judgements of such collaborations as ‘good’ or ‘bad’. We expand on each of these at greater length below. However, it is worth noting here at the outset that while there are important differences and potential tensions between these criteria for our interviewees,they are also seen as interconnected in the making of their overall assessment of whether a collaboration was deemed to be ‘good’, where the term ‘good’ is seen to encompass both instrumental and intrinsic considerations. 1. Active involvement in cutting-edge, interesting science It was clear from all of those interviewedthat the perceived quality and social value of the sci- ence being proposed was an important factor in their decisions about whether or not to join a collaboration. Whilst this was necessary for good collaboration it was rarely sufficient. A signif- icant additional criterion for the researchers we spoke to in their assessment was whether the proposed collaboration offered them an opportunity to be actively involved personally in cut- ting-edge, interesting and outstanding science. For researchers in low-income settings, worries of this kind were common and often grounded in personal experience, or in that of colleagues, of having been relegated to the role of “a glorified field worker” (East African Principal Investigator, East Africa). That is, of being seen as responsible for providing samples but being excluded from the creative, interesting and ‘scientific’ features of the collaboration. As the following quote, from another African Principal Investigator illustrates, worries of this kind can be a major factor in deciding whether or not a collaboration is worth joining: . . .one of the things that kind of makes me a feel a bit uncomfortable about joining . . .is when you feel that [collaborations] are really about getting access to samples without really having Good and Bad Research Collaborations PLOS ONE | DOI:10.1371/journal.pone.0163579 October 13, 2016 4 / 19 any kind of contribution [. . .] And I think that’s something that I’d be wary about, where I felt amI really going to contribute anything, you know, intellectually? (African Principal Investigator, East Africa) As this quote illustrates, where it is believed that invitations to join collaborations are judged to be primarily for access to samples or motivated by tokenistic reasons—such as to give the appearance to funders that researchers from a particular region were involved—these were con- sidered ‘bad’ collaborations by those researchers we interviewed.An additional implication of this worry was that, in instances where researchers felt that collaborations were too large, or central- ised for them to have any realistic prospect of making a meaningful contribution or in cases where they believed for other reasons they were being invited to participate for reasons unrelated to their scientific contribution, researchers were likely to refuse or to participate reluctantly. All the researchers interviewedvalued being included in creative science and the potential for involvement of this type was foremost in their minds when evaluating the merits of collaboration. 2. Effective Leadership Notwithstanding the importance of active personal involvement in the setting of scientific agendas, great emphasis was also placed on the value of effective overarching leadership of the collaborative project or network. Generally, in our interviews, this took the form of an acknowledgement of the key role played by an individual with overall vision for the collabora- tion or international network. The involvement of an effective leader was also a common rea- son given for accepting an invitation to be involved in collaboration. A good leader was someone highly respected in their field and someone with an impressive scientific vision: I have to say she was always way ahead of us in her thinking in the area but we have always admired her. . . (American Principal Investigator, East Africa) It was also clear, that in addition to his or her intellectual leadership the leader had to be someone who was genuinely interested in the relationships underpinning the collaboration. This was seen as very important in developing trust among collaborators. The following quote illustrates the complementarity of these two characteristics: Box 1. Criteria for used by researchers in assessing collaborations • Active involvement in cutting-edge, interesting science • Effective Leadership • Competence in and commitment to good scientific practice • Capacity building • Respect for the needs, interests and agendas of all partners • Opportunities for discussion and disagreement • Trust and confidence • Justice and fairness in collaboration Good and Bad Research Collaborations PLOS ONE | DOI:10.1371/journal.pone.0163579 October 13, 2016 5 / 19 Yeah, clear leadership, he provides it, everyone else follows which is. . .The best form of gover- nance is benevolent dictatorship and I think that's probably slightly controversial but I think that's what [is required] they trust him, he evidently cares. . .he cares about that relationship and the network. . . I think that's probably one of the reasons why it has worked so well. (European Network Coordinator, Europe) What this quote suggests is that where the collaboration is led by an effective leader, who is strongly committed to the maintaining of trusting, respectful relationships, this need not be in tension with—and may in fact be essential for ensuring—the active involvement of partners in resource-poor settings in interesting, cutting-edge science. In addition to the importance of overarching network leadership, effective leadership was also seen as an important characteristic at individual partner institutions, for example where these were under consideration as potential partners. This clinical trials manager suggests that: . . .[a good] site, a good collaborator, or a good collaborator with his team, or her team, will be somebody with good leadership skills and can set up a team that can deliver. (Asian Clinical Trials Manager, South East Asia) This informant argues that a good collaborator is someone who can ‘deliver’ and this is seen to call for someone with good leadership skills. Such judgements were not always or only about particular individuals. Notions of a good institution, research team and leader were often used interchangeably and underpinning such assessments was the ability to deliver good data in a timely manner. In summary, when making assessments about a potential or current collaboration, whether from the perspective of a researcher in a low-income country being invited to participate or from that of a senior researcher considering potential partners, the type of leader involved was deemed important. Those interviewedemphasised the importance of leaders who were engaged in and making, a valuable contribution to their subject area. In addition, a good leader was interested in and cared about the details of the collaboration and ultimately is someone capable of ensuring the timely delivery of good data. It is important to note that the judgement being made here is one that has both a scientific and an ethical or moral component. 3. Competence in and commitment to good scientific practice Given the importance placed on scientific value and good leadership, and their close connec- tion to the concept of a person or institution that is able to ‘deliver’, it is perhaps unsurprising that researchers also emphasised the scientific competence of potential collaborators and their perceived commitment to good scientific practice as key factors in judging whether a collabora- tion or collaborator would be likely to be a good or bad one. It is also unsurprising that in their assessments of competence and commitment to scientific practice researchers tended to emphasise ‘deliverability’ and ‘timeliness’ as signs of a good collaborator and collaboration. Our intervieweesargued that good collaborators were those who were, or were likely to be, able to deliver data, samples and analysis in a timely way to enable research to proceed smoothly. In the quote below, for example, a clinical trials manager describes having entered into a collaboration agreement with a research team only to find that the collaborator had not been able to meet their expected recruitment timelines: Interviewee: They [. . .] got their approvals later than anyone else and once they got approval they recruited one patient and that was about 9 months ago. Interviewer: And how many were you hoping for? Good and Bad Research Collaborations PLOS ONE | DOI:10.1371/journal.pone.0163579 October 13, 2016 6 / 19 Interviewee: 120. (Asian Clinical Trials Manager, South East Asia) After giving this example, the trials manager went on to elaborate on the factors informing their view of a good collaborator, contrasting the group mentioned in the quote above with another who was judged to be a ‘good’ collaborator despite working under difficult conditions: . . .a good collaborator, a good team, an example would be you know the [..] site that I told you about. . .now that’s a good collaborative site because the team is really good, they have the patients, malaria is the problem for them. So although it’s very difficult to get to and they are in [a distant location], we like themasa site becausethey’re veryproductive, theydeliver. (Asian Clinical Trials Manager, South East Asia) It is worth noting that the term ‘good’ is being used in the quote above, and in other inter- views, in both instrumental and moral senses. Instrumentally, the ‘good’ collaborator is pro- ductive and delivers data. Here the concern is about good scientific practice during a collaboration. Morally, a collaborator’s ability to collect data: “[. . .]in a structured way and enter[ed] into a computer system and analyse[d]. . .” (European Principal Investigator, Europe) indicated that they respected their role and the time of colleagues. Similarly, it was a bad collab- orator in both senses—scientificand moral—who sent samples in “inappropriate packaging, not cool, melted, [with] lids. . .off”(European Laboratory Manager, Europe) because this meant a huge amount of salvage work and achieving research aims more difficult.Hence, having sam- ples—and data—delivered on time and in a condition which followed procedures indicated that the collaboration’s aims and objectives were being respected and that ultimately an indi- vidual or team was good, in the sense that they took their responsibilities seriously and respected colleagues. The use of ‘ability to deliver’ and ‘compliance with standard procedures’ as criteria for a good collaborator was not seen as unproblematic, however, particularly, but not exclusively, by researchers in low-income settings. This was because this assessment often omitted consider- ation of real world challenges in conducting research. These included significant external con- straints which meant that even highly committed and responsible researchers were likely to experience difficultiesin following procedures. As an illustrative example, a PI working in a low-income setting describes how practical and logistical concerns shape the timely delivery of data and its condition in at the point of arrival at a biorepository. She explains that: . . .when you operate from Africa you don’t control the time where you sample has to go because [. . .] the few flights that are available, not all of them would accept certain a type of material [e.g] DNA. We’ve been trying to ship this thing for nearly 45 days and in the morn- ing the guy calls and says, “Hey I found a spot on [the next flight]. If you can get your samples ready in an hour it will go today, otherwise we’re going to have to wait one week.” [. . .] so the guy came . . . he saw us we were frantically getting the things together. . .but the goods were gone that day. (African Principal Investigator, West Africa) In situations where samples have not have been delivered according to procedure or are late, there can be external mitigating factors which have meant that compromises have had to be made between timeliness and orderliness. It noteworthy that these factors were widely acknowledged by the research actors we interviewed,wherever they came from. For this reason, many researchers mentioned that they did not make assessments of their collaborators in low- Good and Bad Research Collaborations PLOS ONE | DOI:10.1371/journal.pone.0163579 October 13, 2016 7 / 19 income contexts based solely on deliverability and timeliness. These judgements were multifac- eted. It was clear however that having ‘the right attitude’ was very important for all involved. Worries about this were expressed in a variety of ways. For example, one researcher, a labora- tory manager in a high-income country, who was very aware of and understanding about the challenges facing researchers in low-income settings, was nonetheless very unhappy about, [being treated like] they're doing you a favour by having sent you what they've sent you. . . (European Laboratory Manager, Europe) despite the significant increased workload this caused for those receiving and curating the samples. 4. Capacity building In addition to the opportunities it might provide for personal involvement in cutting edge sci- ence, senior researchers in low-income settings also placed a great deal of emphasis on the importance of ‘capacity building’ for others as a key component of a good collaboration. For all of those we interviewed,capacity building was considered crucial, and its absence in collabora- tions viewed very negatively indeed. The term ‘capacity building’ was employed in different ways to refer to activities including training courses or fellowship roles for early-career researchers, studentship funding opportunities for PhD students, and the provision of scientific equipment …