“A symbiosis in the sense of improving patient benefit”

"Implementation science and health services research must converge even more closely at their intersection and collaborate more closely than before." This is the plea of Prof. Dr. Juliane Köberlein-Neu, Chair of Health Services Research and Health Economic Evaluation at the University of Wuppertal and Congress President of the 23rd German Congress for Health Services Research. In an interview with Monitor Gesundheitsforschung, Professor Köberlein-Neu explains the conference title, "Implementation Knowledge Creates Innovative Care," and discusses what she hopes to achieve in the future for both disciplines.

>> Many experts assume that we don't have a problem of knowledge, but rather one of implementation. Your conference, which you chair as the conference president, is entitled "Implementation Knowledge Creates Innovative Care." What is the state of implementation knowledge in Germany?
This cannot be quantified. However, it is observable that the topic is increasingly gaining political and scientific attention. Especially in the field of health services research, more and more researchers are interested in how they can integrate implementation issues into their projects; they are increasingly engaging with the topic and building implicit knowledge about it. Other countries, such as the USA, Great Britain, Canada, the Netherlands, and Switzerland, are now conducting more systematic research on implementation and making the findings explicitly available for research and practice.

Do you see a strong focus on implementation in this country?
I wouldn't call it strong at this point. Nevertheless, the effort to integrate initial implementation questions into the usually predefined project framework is evident, such as barriers and facilitators of implementation or success factors in the implementation process. However, stand-alone implementation studies have not yet been fully integrated into the funding programs of the majority of funding institutions.
programs because the focus in health services research is primarily on the everyday effectiveness of complex interventions, such as new care models. Therefore, the topic of implementation – at least in my perception – does not yet have a very high priority in the practical implementation of research projects in this country, since the implementation science questions and the associated methodological requirements within research projects often compete with the usual questions of health services research. In other countries – the USA, Great Britain, and the Netherlands are just two examples – there are very specific projects that are carried out solely on the topic of implementation, where the focus is no longer on the everyday effectiveness of a new form of care, but rather on the effectiveness of strategies for establishing it.

Is Germany lagging behind?
I wouldn't say so. In some countries, implementation research is simply conducted much more explicitly than here. If we look at publications from Germany, we see that the majority of publications on funded projects address, on the one hand, the effectiveness of new forms of care under everyday conditions, but on the other hand also describe the results of process evaluations that have revealed barriers and facilitators to the implementation of new forms of care, information about the target groups reached, or findings on implementation fidelity. However, they often stop at this point.

What is the goal of implementation research?
Building implementation knowledge means examining questions that go beyond the aspects just described. These include questions that ask which implementation strategies were used and how effective they were in specific application contexts; which barriers the implementation strategies actually address or which facilitating factors supported them; what costs the implementation process of a new care concept incurred from the perspective of those implementing them, and so on. A very concrete example: When a new care concept is implemented to improve medication therapy safety, it is not only of interest which barriers arose during implementation and which facilitating factors facilitated it. This information certainly helps in planning further initiatives on the topic. Even more relevant, however, would be the knowledge of whether a specific implementation strategy, for example, "external support," was actually able to remove the barrier(s) for which it was developed. I think there is a lot of tacit knowledge in Germany about various aspects of implementation. It is often not sufficiently documented on the publications page—and I personally do not want to exclude myself from this.

Isn’t that also due to the way funding is provided in Germany?
That is certainly true. Often, the time and financial resources available for a single project are insufficient to address both the classic questions of health services research and specific implementation science aspects in equal measure. From a methodological perspective, however, this is not even possible, as each area has its own requirements. Perhaps, however, there has been a lack of human resources, particularly among junior researchers, to bring the mainstream of implementation issues, such as the focus on the effectiveness of implementation strategies, onto the research agenda. In Germany, explicit implementation science training, continuing education, and continuing education programs are still in the development stage. Currently, there are only a few study programs.
processes that either explicitly address implementation science in
They have degrees – as is the case in Heidelberg, for example – or teach implementation science in depth. Training, continuing education, and continuing education opportunities for so-called implementation practice are almost nonexistent in Germany.

Wouldn’t a different support culture be needed to achieve better implementation insights?
In my opinion, the funding culture in Germany is currently undergoing a period of upheaval. Funding agencies are also increasingly recognizing that, without intensively addressing the topic of implementation, we are wasting research resources because valuable scientific findings are not being put into practice. For example, let's look at how the Innovation Fund has developed in recent years. Its current design clearly demonstrates the effort to provide a funding framework for bridging innovations with proven everyday effectiveness into healthcare practice. But that alone is not enough, because all funding bodies depend on which applications are actually submitted, which reviewers assess them, and their awareness of, for example, the need for certain implementation science questions. What makes things even more difficult is that there are many questions in implementation science that appear to be very application-oriented, but which, from an implementation science perspective, would be described as basic research, such as those on the effectiveness of implementation strategies and their mechanisms of action.

What's the problem? There are grants for basic research.
Funding bodies such as the German Research Foundation (DFG), which places great emphasis on basic research, often classify implementation science projects as application-oriented research rather than basic research in the true sense. What funding institutions call or consider basic research therefore clashes with how research is conducted in implementation science and with the methodological breadth that can be applied. But I think this will also increasingly change in the coming years.

Where can implementation science be implicitly funded? Through the Federal Ministry of Education and Research (BMBF), the Federal Ministry of Health (BMG), or the Innovation Fund, within whose framework quite a few implementation-related questions almost regularly arise?
All of the funding institutions mentioned should have a strong interest in advancing implementation science issues, and I believe they do. However, greater awareness must be raised that the questions addressed in the projects may be different from those previously addressed—namely, not just the pure everyday effectiveness of a new form of care and the barriers and facilitating factors observed during its implementation. It is also important that projects may have a different structure and dynamic—for example, they may need to be much more iterative if implementation becomes a more prominent research topic.

Let’s illustrate this with an example.
For example, we have a new health care concept that needs to be evaluated for its effectiveness in everyday care through a study – ideally through a randomized controlled trial. Incidentally, this is by no means always the case in health care research, because we often cannot achieve this level of evidence.

Where is the implementation?
Exactly. This would require either a separate project or at least a clearly defined and funded component within the same project that explicitly addresses implementation issues. This would allow for in-depth questions to be asked and researched, such as which implementation strategies are more or less effective at anchoring the new care concept in question in different contexts of standard care. Ideally, the implementation strategies would also be compared with each other in a control group design – and this "implementation knowledge" cannot be provided solely by a process evaluation, which we often find in projects as a supplement to summative evaluation.

The aim of the Innovation Fund is to support projects that can be integrated into standard care.
This, in turn, requires knowledge about the effectiveness of implementation strategies, which has been researched only to a limited extent in the projects funded to date. In addition, many projects start with certain ideas about implementation strategies that are rarely questioned and are not very context-specific.

Which one?
There are many implementation strategies. For them to be effective, they must be specifically tailored to the barriers and facilitators expected in an implementation initiative or that actually arise during the process. However, it is not uncommon for a project to begin with a scattergun approach, with training courses being scheduled for nursing staff, doctors, or other healthcare professionals involved in the introduction of a new care concept—because training is always part of the process. Supplementary information materials, guidelines, or recommended actions are often included. However, these individual measures only have added value if they specifically address barriers and facilitators to implementation that actually exist at a facility participating in the project. Of course, during the application process, one can initially develop a specific spectrum of expected barriers and facilitators based on literature; after all, one must have a starting point for budget calculations. However, it should be clear that a project cannot begin with a fixed set of measures developed without a specific context; rather, projects must be designed more dynamically. It would be useful, for example, to explore the actual barriers that arise during the implementation or transfer into the reality of care in a specific context before or right at the start of a project. Once these barriers are identified—for example, a lack of support from management or organizational values that conflict with those of the new measure—it is much easier to define appropriate strategies. Once this has been done, a study can evaluate various options for influencing a barrier and then test these against each other. This is precisely a core task of implementation science, whose object of investigation is not the actual care concept, but the question of how an already evidence-based care concept can best be transferred into practice.

Best means?
Which strategies work most effectively and efficiently during implementation? The strategies identified and evaluated in this way can then be useful not just for one, but for several new forms of care. Therefore, it is important to research implementation strategies in such a way that generalizable knowledge can be derived from the results. In all of this, however, it is important to keep in mind that implementation strategies only create meaning and benefit in a very specific implementation context, namely when the barriers and facilitators they address are present.

In the meantime, many pilot projects have been supported by the Innovation Fund.
They will likely rely on similar implementation strategies, but without having previously evaluated which ones are best in their respective contexts. Would it be advisable for health services research or implementation science to create a methodological and theoretical toolbox for the most optimal implementation strategy, for example, for pilot projects?
Let's call the whole thing a bundle of strategies, because implementation is always context-specific and usually requires addressing multiple barriers. More funding projects would need to be initiated for such barriers. Staying with the topic of pilots, follow-up projects would now need to determine which implementation strategies work best in which situations and contexts. The projects could, if necessary, build on data already collected from previous projects and analyze them from a different perspective or conduct in-depth studies using new data.

Context-specific work seems to be what is special in the field of implementation science.
Exactly. An in-depth analysis of the context is often an important starting point in implementation projects – both from a practical and a scientific perspective. But that also means that the whole thing is becoming increasingly complex. Let's take a certain barrier, "knowledge," that stands in the way of the introduction of a new form of care. At a higher level, this barrier may have the same name for different locations, but at one or more locations it may develop very individually in interaction with the respective context. For example, it may be that at one location the "training" strategy works better for this barrier, while at another location the "information material" strategy works better. This is simply because every barrier is different in each context and therefore needs to be addressed specifically. The challenges of implementation science lie in generating context-specific knowledge on the one hand, while at the same time achieving a certain degree of generalisability of its findings on the other.

Can this even be achieved in “normal” health services research projects?
It's just very difficult, but it doesn't have to be. What's more important is that we develop an awareness that the evidence-based design of implementation processes requires additional knowledge that may not be able to be generated in a single large project. For example, a project that primarily focuses on everyday effectiveness can always be supplemented by a second one that explicitly addresses how implementation can best be designed afterwards. Only in this way will it be possible to gain the full breadth of knowledge required to successfully integrate new forms of care into standard care. It's clear that every project must use the methods, theories, and study approaches necessary for the fundamental questions. And because the methodological requirements vary depending on the focus of the question, it's very rare that one can generate evidence of the same quality on everyday effectiveness and on implementation in the same study.

Is implementation science taken into account right from the start?
Not everywhere yet. However, I believe that a large portion of health services researchers are already trying to get the best possible results from their projects, including for implementation. The fact that this hasn't yet been achieved comprehensively is simply because there are still too few people who have made implementation their focus, or have been able to make it their focus, in health services research and beyond.

This may also be due to the fact that health services research in Germany has historically developed largely at medical universities and colleges, while implementation science has a sociological character.
That's certainly a possible reason. But I think it's also partly due to the current funding opportunities. In every scientific field, there are advancements that will generally move in the direction permitted by the existing funding opportunities at a given point in time.

In this respect, only what can be achieved within the funded projects will be implemented.
This is the major difference, especially compared to the Anglo-Saxon and Nordic countries, or even our neighbors like the Netherlands: Centers for implementation science and dissemination now exist there. Switzerland, too, has its first institutions dedicated intensively to implementation and conducting extensive research in this area.

Let's discuss implementation science and health services research at this point. Aren't they two sides of the same coin?
For me, they each have their own unique character, but they share a large overlap due to an overarching goal that can unite both and the shared range of methods they use. Implementation science should be understood as a cross-sectional area used by many disciplines, such as the education sector, which focuses on how innovations can be introduced into the school system, for example. Implementation science has developed a considerable amount of basic knowledge in this area in recent years.

Does something similar apply to the field of health services research?
A person who explicitly sets out to generate implementation science findings has a different research interest than someone in "traditional" health services research. However, both work toward the same goal: improving healthcare. That's why it's so important to understand what and how research is conducted in both fields, as well as to reflect on and consciously position yourself within this field.

Let us illustrate this again with the example of the pilots.
For example, a traditional health services research project would ask whether a new form of care that included guides led to improvements in certain patient-relevant endpoints in everyday care compared to current standard care. An accompanying process evaluation would also examine the extent to which the new care concept was implemented in everyday care, or which components of the new care concept were provided, in what form, and with what quality. Furthermore, it would likely investigate which patients were actually reached by the new form of care.

These are classic questions in health services research.
Exactly. From a more implementation science perspective, the same project would investigate which implementation strategies were used and developed, and especially how these addressed and changed existing barriers.

Although, as you already pointed out, these often emerged ex ante, almost in a vacuum. Or were simply implemented because someone said, based on gut feeling, that training is always needed.
From an implementation science perspective, one would first analyze whether a barrier that can be addressed through "training" actually exists. For example, if there is no real knowledge problem, an implementation strategy focused on "training" will provide little or no benefit. Implementation science, on the other hand, also analyzes, for example, which implementation strategies were used as planned for a specific context and which success steps in the implementation process exist within an organization.

The focus is therefore on implementation activities and no longer on the new form of care itself.
Quite right. I also want to make it clear that traditional process evaluation, as we conduct it in health services research, is not the same as implementation science research. Sure, there's some overlap, but there's no complete congruence.

Would you personally consider yourself an implementation scientist?
I don't know if anyone would call themselves a pure implementation scientist. We all have our own academic background and background. Many people who practice implementation science come from psychology or sociology. I wouldn't describe myself as an implementation scientist in that sense, but rather as someone who genuinely strives to pose and answer questions from implementation science within health services research. To do so, however, I have to constantly remind myself that from this perspective, some questions need to be formulated and researched differently.

Implementation scientists are also relatively rare in Germany. Examples that immediately come to mind are Professor Wensing from Heidelberg, Professor Vollmar from Bochum, and Professor Roes from Witten/Herdecke, who will likely attend the conference you will chair as conference president at the end of September this year. Do the topics just touched upon and discussed resonate with them? If so, in which of the four strands?
This time, we tried to define the strands a little differently and build them on one another. They follow the process by which knowledge is transferred into healthcare practice.

The program committee and you have defined four strands for the German Congress on Health Services Research (DKVF), which will take place in September. These are: Strand 1: Creating Knowledge, Strand 2: Presenting and Interpreting Knowledge, Strand 3: Disseminating and Communicating Knowledge, and Strand 4: Applying Knowledge in Practice. Strand 1, for example, will focus on the (further) development of concepts, theories, and methods. What can we expect there?
The first strand is not so much about the findings that studies generate, but rather about the question of which methods can be used. For example, what alternatives there are to randomized controlled trials, such as the area of realist methods and their associated potential for evaluating complex interventions. Methods for dealing with healthcare-related data, as well as mixed-methods approaches, are also discussed.

What can congress attendees expect in the second strand?
The topic of "presenting and interpreting knowledge." This focuses on the implications that can be derived from the results obtained. "What happens now with it and why?" – this is the core question of implementation science. Consequently, the third strand is about how the acquired knowledge can be disseminated and communicated. This involves building competence, capacity building in both a research methodological and practical sense. And, of course, vice versa: What can we learn from practice and what can we incorporate into our own research in the appropriate form.

And in the fourth strand?
The fourth and final strand is about reflecting on project experiences and applying them in practice. Barriers will also be discussed here: what barriers actually exist and how they can be influenced. And where we may simply be limited. I would also like to particularly highlight the Patient Day, which will take place on Thursday parallel to the rest of the scientific program. Together we want to explore how participatory research can be even more intensive in the future.

If you are interested in the conference topic “Implementation knowledge creates
innovative care” in a nutshell, then...
...implementation science and health services research must come together at their intersection even more closely and collaborate more than before. Both disciplines must manage to use the symbiosis they can enter into even more effectively in the interests of improving patient benefit. Both – health services research on the one hand and implementation science on the other – can benefit enormously from one another. This begins with the fact that in every project the exact questions that will ultimately be needed for implementation in everyday health care must be asked and then researched as early as possible. Implementation science has a whole spectrum of theoretical approaches for this purpose. It would be a tremendous win-win situation if we could bring the two fields even closer together.

I'm really looking forward to this conference. See you there, and thank you for the conversation.

The interview was conducted by MVF editor-in-chief Peter Stegmaier.