Article Text
Statistics from Altmetric.com
In 1999, the Institute of Medicine (IOM) unveiled the dire need to make healthcare safer.1 In an effort to reduce harm, improve performance and minimise cost, quality improvement (QI) methodology was identified as an ideal approach to closing the quality gap.1–3 In the years that followed, the dissemination of publications using QI methods increased significantly.4 By 2008, the first edition of the Standards for Quality Improvement Reporting Excellence (SQUIRE) Guidelines was published in an effort to support the breadth, usability and rigour of scholarly healthcare improvement work.5 Seven years later, the modified SQUIRE 2.0 Guidelines bolstered the thorough, theory-driven reporting of interventions and improvement efforts to spread generalisable and actionable knowledge.6 Furthermore, the advent of learning health systems, defined by the IOM as a system ‘designed to generate and apply the best evidence for the collaborative healthcare choices for each patient and provider; to drive the process of discovery as a natural outgrowth of patient care; and to ensure innovation, quality, safety, and value in health care’, has brought the value of healthcare improvement work to the forefront in academic centres.7 8
Incorporating continuous improvement and data-driven learning with traditional research methodology and routine care delivery is now common and encouraged. Consequently, the difference between classic human subject research (traditionally separate from routine clinical care) and QI research is increasingly blurred. Newer scientific fields such as implementation science focus on how to implement and disseminate evidence-based practices.9 QI research uses a different framework of improvement science to understand the effects of an intervention on important quality problems.10 While these related fields focus on what gets done in clinical care (compared with what is known), their definitions may differ at local, national and international levels. Regardless, the fact remains that scholarly dissemination of publications using research methodologies that do not fit into the traditional clinical research model is rapidly expanding. Now is a pivotal time to evaluate the application of authorship criteria to these fields—a consideration that is missing from SQUIRE 2.0.
Elucidating similarities and differences between QI and implementation research is beyond the scope of and not relevant to our argument. Instead, we argue that these related fields warrant a unique approach to authorship determination, which has not been adequately addressed. We discuss how issues around the generalisability of the knowledge being sought, team composition and timing of authorship determination can impact who meets authorship criteria in QI and implementation research. We conclude by offering a series of questions research and project team leaders can ask themselves to assess authorship more fairly and appropriately.
Authorship in research
Guidelines for determining scientific authorship have existed for over 30 years, rooted in practices of traditional human subject research.11 Peer-review journals have adopted these guidelines to apply to all published work, including QI and implementation research. While there has been some muted criticism and proposed adjustments, little has changed since the guidelines’ inception.12 Most peer-review journals assert that they abide by the International Committee of Medical Journal Editors (ICMJE) guidelines, which necessitate that an author meet four criteria: (1) the author made substantial contributions to the conception or design of the work or the acquisition, analysis or interpretation of data; (2) drafted or revised the work critically for important intellectual content; (3) gave final approval of the version to be published; and (4) agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.13
In the ICMJE guidelines, the subjective determination of ‘substantial contribution’ is the most ambiguous. Is it determined by the time spent on the project or the value of the individual’s efforts? How is the role held in the project’s design and implementation weighed? When is the decision made that an author met ‘substantial contribution’ criteria? Who makes that decision? While the answers to these questions have been perseverated over for traditional human subject research, less attention has been given to answer these questions in QI and implementation research.14–16 We believe the following specific factors should be considered for authorship determination in QI and implementation research: generalisable knowledge; team composition; and timing of authorship determination.
Generalisable knowledge
Early in the project’s design, team leaders should consider if the proposed study has the potential to contribute to generalisable knowledge. Generalisable knowledge may be shared through publication. However, not all projects result in or are intended to produce generalisable knowledge. QI projects use improvement science methods, such as the Model for Improvement and plan, do, study and act (PDSA) cycles, to test a series of small interventions aimed at systematically implementing knowledge and achieving immediate, local, care improvements.3 17 Similarly, implementation science projects use methods to incorporate evidence-based interventions into practice.9 Sometimes the implementation of an intervention, such as central-line care maintenance, may be novel to an investigator’s institution but already well described in the literature in a variety of care settings. In such cases, the implementation work may be less generalisable. Conversely, an investigator who is the first to implement an antibiotic stewardship programme in a rural paediatric clinic may disseminate this work for others to adopt. The generalisable knowledge spread through publication of these project types relates to the success of the intervention in improving healthcare or the novelty of the intervention itself (which often times has already been studied) and to the context in which the intervention was developed and evaluated. In fact, SQUIRE 2.0 highlights the importance of thoroughly describing local context when disseminating knowledge gained, so readers can determine if this knowledge is applicable to their own local context.6 When designing a study, team leaders should attempt to a priori investigate if their research fills a knowledge gap related to context and/or novelty of the intervention. This may result both in a more robust and informed project, as well as help team leaders determine if the knowledge gained is generalisable and therefore publishable.
Team composition
QI and implemention research focus their efforts on the care delivery process. Project teams tend to be large in number and scope of experiences, particularly compared with traditional research teams. Traditional research teams may include a primary investigator, statistician and research or laboratory assistants whose roles are limited to their field of expertise within the project. QI and implementation research team members come from the pool of stakeholders impacted by the project. For example, a QI team focused on improving a hospital’s discharge medication reconciliation process may include: a team leader (primary investigator), statistician or data analyst, QI specialist, executive sponsor and several expert representatives from the different processes targeted by the project, such as attending physicians from various subspecialties, resident physicians, nurses, pharmacists and patients and/or family members. Partnerships between researchers and patients and/or family members is crucial and supported in the literature.18 19 Patient’s and family member’s unique perspectives and contributions deserve recognition, as they inform planning, implementing and adapting interventions in complex healthcare systems. The patient’s presence alone drives different behaviours and innovations on a QI team and so the opportunity of authorship should be supported.18 19
The project team’s composition consists of many people who vary in how they substantially contribute to the research design, implementation efforts, data acquisition, analysis and interpretation. A similar team-based approach to determine substantial contributions towards authorship is warranted. Teams should collaboratively discuss the general criteria for ‘substantial contribution’ based on each member’s specific role and responsibilities. Contributor Roles Taxonomy provides a starting point for teams to classify their diverse roles, which may be translated to the authors’ contribution statement in the published work.20 Discussing what constitutes ‘substantial contribution’ should be done at the beginning of the project and revisited on an ongoing basis, so all members stay informed. While the definition of substantial contribution will depend heavily on the nature of the project, some general criteria could be based on leadership, project design, crucial contributions to tests of change or implementation involvement. These contributions may be measured based on amount of time invested or number of PDSA cycles involved. Iterative, team-based discussions with clear, measurable criteria for the definition of ‘substantial contribution’ should be routine practices for any QI or implementation team. Team leaders should encourage and support team members who lack experience or comfort with scientific writing to participate in this part of the research process so authorship criteria can be met. However, those individuals who meet the team’s standards for substantial contribution but do not wish to participate in manuscript writing and/or review should still be recognised for their valuable contribution with an acknowledgement.
Timing of authorship determination
Because of the iterative, adaptive nature of QI and implementation projects, team members vary in their roles and responsibilities throughout the projects’ duration. Although it is important to consider the intent to disseminate knowledge at the projects’ inception so team members’ contributions can be regularly and objectively assessed, often team leaders cannot reliably identify who substantially contributes at the beginning of a project. QI and implementation projects often have multiple, changing cycles where ideas put forth by the project’s team are constantly developed and tested. As previously described, these teams are diverse in their composition. Depending on the direction the project takes, team members who were engaged initially may no longer be as engaged after their related tests of change are maximised. Similarly, new tests of change may necessitate new team members who meaningfully contribute, even at a very late stage. For example, technology-based changes in an electronic health record often occur after proof of concept tests from paper-based interventions. The informatics team may eventually substantially contribute to the generalisable knowledge developed from the project but may not have been present at the projects inception. While the factors that determine substantial contribution should be regularly discussed at team meetings, delaying the determination of whether an individual made a ‘substantial contribution’ to the project’s constantly evolving design may be more appropriate after the study period ended and before the writing phase begins. Team members who substantially contributed may deserve recognition through authorship, even if their contributions only spanned a small part of the project’s timeline.
Considerations for QI and implementation research authorship
With submission of QI and implementation research projects for publication as prolific as ever, team leaders should ask themselves three questions when approaching the determination of authorship for manuscript publication (table 1).
Is there a chance this work will yield knowledge that is generalisable enough to warrant dissemination through publication? An answer of ‘yes’ should trigger team leaders to a priori set up a plan for a team-based approach to assessing members’ roles and contributions at regular time points throughout the project’s duration. If the answer is ‘no’, team leaders should consider keeping track of members’ contributions individually, as often, QI and implementation projects identify knowledge to be disseminated, although it was not the initial project’s intent.
What does substantial contribution mean to me and my team? The approach to authorship determination should be no less rigorous than any other aspect of the QI or implementation project. These projects require consistent attention to detail, iterative methodologies and constant reassessment. These same standards should be applied to establishing criteria for substantial contribution. This should be a team-based discussion that is started early in the project’s design and addressed regularly as it is carried out. General criteria that are measurable should be established and tracked, so as not to disadvantage team members whose contributions during the earlier project phases were more significant. Team members who meet criteria for substantial contribution but do not want to fulfil the other ICMJE criteria can still be recognised through an acknowledgement.
Can the determination of authorship wait until the end of the project? By having clear framework for determining substantial contribution and delaying the decision of authorship until the end of the project, team leaders can more appropriately apply the ICMJE authorship criteria. Team members who become eligible for authorship by making ‘substantial contributions’ throughout the project’s duration should be specifically asked if they commit to: (1) revising the written manuscript; (2) approving the final version; and (3) accept responsibility for the work’s accuracy and integrity—the three remaining ICMJE criteria.13 Team members who have substantially contributed can choose if they want to continue with the responsibilities of writing and/or revising a manuscript towards publication by providing meaningful comments after a thorough draft review. Then, they can approve the final version and accept responsibility for the published work.
It is important to note that project leaders should not feel despondent if the first time they think about publication is not until the project’s later stage. In particular, this is likely to be a common experience for many QI teams. While we suggest starting to consider authorship early, our approach can still be used to help guide authorship determination at any stage of the process. Considering the factors outlined above will help advance the growing body of QI and implementation research literature and broaden the opportunities for authorship to previously excluded team members.
Ethics statements
Patient consent for publication
References
Footnotes
Twitter @philips_kaitlyn
Contributors All authors conceptualised this essay and substantially contributed to the design of the publication. KP drafted the initial essay, and MLR and EC critically revised it. All authors agreed to the final essay prior to submission.
Funding This study was funded by National Center for Advancing Translational Sciences (Grant number: UL1TR001073) and Agency for Healthcare Research and Quality (Grant number: K12HS026396).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.