With an entire book devoted to exploring collaborations in drug discovery and development, there seems to be little question as to the value of such collaboration. There is a widespread call for interdisciplinary research as the key to future advances in research [1–3], and consortia between various public and private organizations have been advocated as key in addressing problems of technology and drug development [4, 5]. That call extends to data sharing that transcends the traditional intellectual property protection culture of companies and even universities [6]. Such precompetitive collaboration as advocated for biomedical research has proven precedence in Sematech, a research and development consortium formed by semiconductor firms and the U.S. government, with the purpose of improving semiconductor manufacturing technology [7]. Sematech served to pool semiconductor R&D for member companies and was shown to reduce the R&D spending of individual member companies, indicating an R&D efficiency achieved by the consortium. Thus, the intuition that pooling of expertise, experience, and resources from various otherwise competing groups can achieve efficiencies beneficial to all members can be objectively confirmed. Furthermore, recent consortia exploring toxicogenomics [4], preclinical and clinical biomarkers [8–10], gene and protein annotation [11, 12], and pharmacogenetics [13] have also shown the power of collaboration in addressing complex problems.

Challenges

There are real issues and difficulties in establishing and maintaining large-scale cross-function and cross-organizational research collaborations. First and foremost is the illogical nature of cooperation. Given that competition between individuals within a species is a basic premise of evolution, its antithesis, cooperation, has been the subject of question for several decades [14, 15]. While much is made of market competition between private R&D companies, nonprofits must also take a competitive view in the face of limited funding [16]. Competition is not just a fact, but a norm, in academic science and in the training of scientists [17, 18]. Welsh and coworkers note that cross-disciplinary and collaborative research are not fostered by university cultures focused on “individual merit” with a “career hierarchy which fosters competition between individuals and between universities” [19]. Funding, academic positions, and influence are all dependent upon a scientist’s ability to excel beyond colleagues in the same field of research, with metrics usually being that of publication priority and number. With first-author status as a prime component of the publication metric, issues of authorship and author order become problematic in research projects that demand a group effort [20–23]. Graduate students and postdoctoral fellows are trained in this environment and are keenly aware that their future careers depend on their ability to compete in it; as such, they rarely acquire collaboration and team skills. So despite the current allure of promoting collaborative efforts, the reality is that human nature, scientific training, and reward structures predispose participants to competition and the pursuit of individual goals.

Collaboration Models

Despite this undercurrent of competition, however, interindividual and interorganizational collaborations have a long history of successes (as well as failures). In fact, a rather large body of research literature on collaboration exists in social science and organizational research. Shrum and coworkers have examined collaborations in the physical sciences [24], but unfortunately, there have been relatively few studies of collaboration in biomedical research settings. One such study examined four multidisciplinary groups and examined forms of collaboration and the influence of personal compatibility, work connections, incentives, and infrastructure [25]. However, social science and organizational work has developed testable models, and it is well worth calling attention to this work and Shrum’s findings concerning scientific collaborations, especially so as even large scientific collaborations are rarely constructed with an eye toward the principles of effective collaboration that come out of this research. It should be noted that there is a parallel body of research concerning effective team construction and activity, which in some cases provides overlapping insight into collaboration models.

Beginning with the work of Wood and Gray [26], theories and models have been advanced to define and analyze collaborations, both successful and unsuccessful. Thomson et al. have accordingly developed a quantitative model based on the definition: “Collaboration is a process in which autonomous or semi-autonomous actors interact through formal and informal negotiation, jointly creating rules and structures governing their relationships and ways to act or decide on the issues that brought them together; it is a process involving shared norms and mutually beneficial interactions” [27]. Their model incorporates five key dimensions: governance, administration, mutuality, norms, and organizational autonomy, with several components of each dimension. Governance entails how participants in collaborations jointly develop rules and roles for interactions and decision making. Thomson et al.’s study affirms the concept that each member in a successful collaboration should have a voice in decisions and planning that is heard by all. The administration dimension includes familiar concepts of roles and responsibilities, communication channels, and implementation procedures. As might be expected, the study supports the importance of clarity, coordination, and agreement on goals. The dimension of organizational autonomy explicitly acknowledges that while there may be collective interests embodied in the collaboration’s goals, member organizations (and individuals) have very real self-interests. A corollary of this inherent tension will be the conflict between any individual’s commitments to the collaboration and those to his/her parent organization. Likewise, there will be tension between the decision-making authorities in the two arenas. Clearly, it is advantageous to the collaboration if member organizations provide participating individuals with a concrete agreement as to the effort they can commit and the authority they have in decisions that affect the member organization. Mutuality, on the other hand, describes the synergistic values of collaboration, the benefits that a member uniquely derives from the whole, as well as the unique benefits that a member brings to the table. Both of these must be clear and real for the collaboration to be valuable. In addition, the mutuality dimension captures the concept of shared interests that members may have and goals that are valuable to, but somewhat beyond those of, each member organization. Finally, the norms dimension captures the concepts of trust, where each collaboration member believes that other partners will meet their obligations.

Both complementary and, in some ways, contradictory conclusions as to the elements of collaboration are reached by Shrum and coworkers in their study of 53 collaborations in physics and related sciences. These authors specifically examine “trust” as a factor and find that while the “trust” that collaborators will have something to contribute is foundational, “trust” on a personal level is not important. Conflict between organizations (or member teams) and between such member teams and an administrative management (e.g., project management) as discussed in this study call to mind the dimensions of governance and administration noted earlier, and echo the tension of those dimensions with organizational (or team) autonomy. Ironically, in these collaborations, those with the least conflict seemed to be those that offered the most autonomy to individual scientists, team and organizational members, and those that did not require participants to share data and analyses. Nonetheless, consensual decision making was viewed as a positive and “successful” element in collaborations, yet hierarchically organized projects were often viewed as successful.

While individual teams, even interdisciplinary ones, are not strictly comparable to interorganizational collaborations, the research on team dynamics can be instructive. Thus, the principles of team assembly that lead to a collaboration network, where there is a dynamic between newcomers and “incumbents” [28], can be considered to be supportive of interorganizational collaborations. Similarly, studies of team leadership can be considered, although as noted earlier, governance and administration of collaborations may function differently from that in a formally empowered team. “Transformational leadership” [29] has been cited as a critical factor in improving the function and outcomes of research and development teams with diverse membership [30]. The four elements of transformational leadership (individualized consideration, intellectual stimulation, inspirational motivation, and idealized influence) would seem to be applicable to collaboration leadership. Yet the governance and administration models for interorganizational collaborations all suggest the importance of member inclusion in developing rules and approaches to decision making. Of equal importance in collaborations is the need for some level of member autonomy. These concerns would not in and of themselves rule out a single point “transformational” collaboration leader, but would suggest that the choice of that leader would best be a collective choice and that the leader would need to be thoughtful of the characteristics of collaborations that are different from those of teams.

Stay updated, free articles. Join our Telegram channel

Join