Enhancing the reporting of implementation research

In 2015, we reviewed and provided a detailed explanation and elaboration of our journal scope [1]. As of 2017, we have no plans to expand further the boundaries of our scope at this point in time. Therefore, our focus remains on the publication of studies examining the implementation of evidence-based healthcare interventions, practices or policies or the de-implementation of those demonstrated to be of low or no clinical benefit or even harmful.

For implementation effectiveness, we seek to publish studies that employ rigorous experimental or quasi-experimental designs regardless of whether they report effects or no effects. By rigorous, we mean those designs that would be eligible for inclusion in the Cochrane EPOC reviews [2]. This can include type 2 or type 3 hybrid designs where there is a dual a priori focus on assessing clinical effectiveness and implementation strategies, [3] but only where there is a clear justification and major element of implementation research. Type 2 designs have dual focus effectiveness and implementation outcomes, here, for example, testing both the effectiveness of brief cognitive behavioural therapy and the implementation strategies [4]. A type 3 is where the primary emphasis is on evaluating implementation, in this instance of a diabetes prevention programme, but where data on clinical outcomes are also collected [5].

We continue to receive a considerable number of studies testing novel clinical or population health interventions, where the effectiveness of the intervention or practice has yet to be established. As our scope focuses on the implementation of interventions of demonstrated effectiveness, we routinely reject these manuscripts (and offer transfer to other BMC journals). These exclusion criteria extend also cover type 1 hybrid designs where the focus is on testing effects of a clinical intervention on relevant outcomes whilst observing and gathering information on implementation [3]. For instance, a clinical trial of primary care management of survivors of sepsis focused on patients’ quality of life as the primary outcome also comprised a range of measures of implementation aspects [6]. Studies of this type fall outside of our journal scope.

Alongside effectiveness, the journal scope also includes economic evaluation and qualitative research that examines different aspect of interventions and context which contribute to effectiveness. This includes the study of adaptation and fidelity, mechanisms of impact and contextual influences on implementation and outcomes, sustainability and scalability as well as the study of influences on provider, patient and organisational behaviour. Crucially, we expect the methods employed in such studies to be an appropriate fit to the question(s) being addressed and be informed by relevant conceptual frameworks.

We also welcome articles that present new methods and articles that question or challenge existing implementation policies, practices, evidence or theory and suggest modifications or alternatives. However, it is worth noting that there is no shortage of frameworks and theories relevant to implementation research [7, 8]. So rather than adding to the current pot, our preference is for empirical studies that build and advance the existing theoretical base. With debate papers, we reject those that fail to ground the central argument within the existing implementation research literature. Many debate papers would be of greater relevance if the arguments posed were based upon systematic reviews of the relevant evidence.

Table 1 presents the types of manuscripts likely to be accepted by or rejected from Implementation Science. This should assist prospective authors to judge whether the journal is a suitable home for their research.

Alongside failure to meet scope requirements, poor scientific quality remains a common reason for rejection. Promoting the development, refinement and quality of implementation research was a key aim of the founding editors [9] and remains so today. We therefore continue to support and promote efforts to improve research quality and transparency.

Implementation Science supports initiatives to improve the reporting of randomised trials. We have adopted the ICMJE recommendation [10] and only normally consider for publication trials that have been registered with an appropriate publicly available trials database prior to enrolment of the first participant/cluster. We will consider retrospectively registered trials on a case by case basis but will require authors to explain the reason(s) for the delayed registration.

Whilst there are no fixed rules about the registration of other study designs, we strongly encourage authors of systematic reviews to prospectively register their review with PROSPERO or other publicly accessible registries.

Over the last decade we have routinely required authors submitting manuscripts that report trials to complete the CONSORT checklist or relevant extension. Similarly, a requirement to complete the PRISMA checklist has been enforced for authors submitting systematic reviews. No other checklists have been routinely or uniformly enforced. As a journal that receives manuscripts covering a wide range of study designs, this has resulted in variation in the standards of reporting of the research that we publish.

Because our aim is to promote research quality and transparency, and as an aid to our readers, reviewers and editors, we now require authors submitting manuscripts (regardless of study design) to complete and include a design appropriate reporting checklist.

The website of the EQUATOR Network provides details of all available reporting guidelines (www.​equator-network.​org). Authors of manuscripts (regardless of study design) should refer to EQUATOR and ensure that they complete and include a design appropriate reporting checklist with their submission. Table 1 includes details of our preferred reporting formats; for those research types where consensus is lacking on reporting format (for example, in qualitative research), we encourage authors to select their preferred checklist.

Improving the quality of intervention description is as much an issue for implementation research as it is for other evaluations of complex interventions. Without sufficient detail, it is difficult for readers to determine what was actually implemented and/or for other researchers to use or replicate the intervention in other studies. Whilst TIDieR has been proposed for use in conjunction with the CONSORT guidelines for trials, [11] improved intervention description is relevant across all evaluative study designs. Other relevant standards for reporting implementation interventions (Standards for Reporting Implementation studies —StaRI) [12] and for reporting behaviour change interventions (Workgroup for Intervention Development and Evaluation Research—WIDER) [13] have been developed and are available. We encourage authors to select their preferred guideline to enhance reporting of interventions.

With all submissions, we expect authors to clearly articulate what is already known and what their work adds to existing knowledge, theory and thinking in the field. Many submissions currently fail to set the work in the context of the existing literature. And so we will continue to reject manuscripts that do not clearly build on current knowledge and understanding or appear to provide limited contributions.

As an open access journal (with open peer review), we are committed to making research and the datasets upon which it is based, publicly accessible. A number of different data sharing approaches have now been adopted across the health and medical literature [14]. At Implementation Science, we have adopted the policies on data availability of our publisher BMC. As part of online article submission, we now ask authors to include an “Availability of Data and Materials” section in their manuscript detailing the conditions by which the data supporting their findings can be accessed. Authors who do not wish to share their data must include a formal statement that data will not be shared and give the reason why. Full details of BMC policies can be found under the Instructions for Authors section of our website.