Research Informatics.org

This past week saw a tremendously productive and invigorating AMIA Spring Congress meeting in Orlando, Florida. During the meeting’s CRI track, a broad variety of topics were covered, from the development of a “human studyome� database, to practical approaches that could be employed to enable the selection and deployment of clinical trial management systems or similar software platforms. A number of cross-cutting themes emerged during the meeting, which I believe will be of interest to readers of this blog, namely:

An emphasis on systems-level approach to CRI: This really was evident in a consistent evolution in thinking from “I think I can� to “I think we can� type project (not quite hypothesis-driven science, but a step in the right direction). In particular, there was a fairly consistent inclusion of socio-technical, workflow, and human factors evaluations relative to the programmatic efforts and study designs that were presented. Perhaps the best news in this area was an even greater emphasis on the development of readily adoptable tools However, despite these positive factors, This transition is still clearly in its early phases

Important work is taking place outside of conventional CRI-centric consortia (CTSA, caBIG, etc.): This was particularly true when looking a the work being presented by international and multinational teams. Most, if not all such efforts were independent or relatively small scale (e.g., single sites, small “grass roots� consortia). There was a strong consensus that current national-scale efforts in the U.S. may be creating silos of translation, rather than enhancing national infrastructures and approaches.

There is a clear movement towards integration of available research information management, analysis, and dissemination platforms into “pipelines�: Such efforts are again largely happening in self-organized collaborative groups motivated by scientific use cases.

And finally, and perhaps more importantly, there was an Increasing maturity shown by CRI researchers in the context of CRI being a distinct discipline of biomedical informatics: However, it was also clear that there is still a need for progress towards a state where CRI researchers not only ask questions, but also answer them.

On a final note, it was exciting to see a strong convergence of thinking, research, and development around a set of common CRI theory and practice themes, as is illustrated in the impromptu figure below.

If you have any thoughts or comments to share about the AMIA 2009 Spring Congress – please share them with your colleagues here!

Unless you have been living in a remote corner of the world for the last several months, you are probably well aware of the current infusion of ARRA (http://www.recovery.gov/) related funds into the national scientific and biomedical research enterprises. In the realm of NIH extramural research, this infusion accounts for 8.2 Billion in additional funds (http://grants.nih.gov/recovery/), divided among the constituent IC’s (institutions and centers) of the institute. Clearly, this is a major catalyst for the national health and life sciences enterprise that has been sorely needed for a number of years (for more details on this issue, take a look at some of my previous postings regarding the grave funding situation facing our community over the past several years). However, the funds are not without strings, which include short time frames in which they must be expended (on the order of 24 months), and stringent reporting requirements. In many ways, the rush by our community to secure such funds for our favored or highest priority projects is reminiscent of the gold rush of 1849 in California, in which a social-epidemic swept the nation, leading to a mass migration and tectonic shift in activities towards a new source of revenue and wealth. Yet, if we open our history textbooks, we also know that by 1859, a short ten years after the discovery of gold in California (especially given the speed of commerce, travel, and communications in the late 1800’s), the combined discovery of silver in Nevada and an ongoing evolution in U.S. monetary policy ended the gold rush. Such a history lesson begs a number of questions, namely:

---

Is a massive infusion of resources into the current biomedical research enterprise, without a broader view towards the evolution of the sector, its practioners, and processes, likely to yield a sustainable model of growth, or set our community up for an even more traumatic evolution in funding and practice in a relatively short time frame (on the order of 24-36 months)?

Will there be a major shift in policy and corresponding social epidemic that will quickly mitigate the cultural, structural, and operational changes we are working so hard to achieve under the current ARRA model, again in a relatively short time frame?

How do we, as a community of researchers, scientists, and practioners, make the optimal use of current ARRA funds and systematic changes being effected by those funds, while balancing our “portfolio� of expertise and initiatives to ensure the continued growth and evolution of our field, especially in light of likely future tectonic shifts in funding and policies?

---

Beyond the preceding questions, there also remains a set of what I would consider similarly compelling question concerning the relationships (or perhaps more conspicuously, lack of systematic relationships) between NIH stimulus related programs and the highly complimentary HITECH act passed as part of the overall stimulus package (http://www.opencongress.org/bill/111-s350/show), especially in the context of the informatics community – a topic that appears worthy of a future posting to this blog!

Weigh in and share your opinion on the questions above, or others that you would like to raise!

[Don’t know what a Forty-Niner is (other than a football team)?: http://en.wikipedia.org/wiki/California_Gold_Rush]

After a lengthy hiatus, I wanted to make a concerted effort to come back to this blog with a posting that was both timely and of potential use to our readers. After thinking about this for a while, I was struck by a number of meetings I have had of late, both internally at OSU, and at the national level, concerned with informatics research and service in the context of large-scale program, project, and center grants (e.g., CCSG, CTSA, P01’s, etc.). As many of you are most likely and acutely aware, informatics research and development occurs with great frequency in the context of such settings, due to the need to provide systematic information management and analysis support to program-specific scientific projects or cores – needs that do not always come associated with a known best practice or solution, thus necessitating both research AND development in order to satisfy end user requirements. This situation presents both significant benefits and significant challenges. First the good news – by performing such research and development in the context of motivating, scientific use cases, the opportunity exists to demonstrate the efficacy, utility, and impact of such informatics solutions in real world settings with results of interest to the broad biomedical community – an outcome that is not always seen in theoretical informatics research (which is often a factor contributing to the limited uptake of such theoretical constructs or methods). Now, the bad news – most funding agencies make it explicit that such research and development within the confines of a largely service-oriented core or program is not necessarily a positive feature, and often if it is emphasized too heavily when applying for such funding, can be a detriment to the overall proposal. Perhaps even more worrisome is that in many cases, the scientific or clinical PI’s of such program/projects share this perspective. This is clearly a sub-optimal situation, and given the current funding environment, it seems we should re-examine our approach to such research programs in order to ensure we can make the most of our limited resources. Therefore, I would propose the following call to action for the biomedical, informatics, and funding communities:

We must recognize that in order to provide truly efficacious and impactful informatics support for scientifically motivated research programs, a combination of both formative research AND service is needed. This fact parallels that bifurcated basic and applied science nature of Biomedical Informatics as a discipline,

Funding agencies need to and should encourage program and project proposals that balance the preceding needs,

Informaticians must work harder to educated scientific, clinical, and funding partners to recognize both the short and long term benefits of such an approach, and

Scientific and clinical investigators must work harder to recognize AND understand the dual-nature of informatics as a basic and applied science, spanning a spectrum from theoretical discoveries to practical services.

However, if we are to realize the preceding goals, we must also work to understand what the optimal balance for research and development versus service in such programs is. In my experience, study sections and other funding review processes outside of the immediate informatics community tend to appreciate a roughly 70%/30% split between service and research efforts respectively. However, such an anecdotal finding is limited at best, and more understanding of this balance is needed to establish prevailing best practices that can be accepted by the entire biomedical community. Creating this understanding will require our community to invest in the meta-analytical process of “Research on Research.� If we can achieve such an understanding of this critical split in efforts, we will be much better positions to make the optimal use of resources in our highest yield and most impactful basic science, clinical, and translational research efforts.

I wanted to take this new posting as an opportunity to comment on the artificial barriers that seems to be defining the pursuit of translational science in general, and more specifically, translational research informatics. So, what do I mean by artificial barriers? Well, if we look at the defining literature and prevailing models in the translational research domain, significant emphasis is placed on the T1 and T2 barriers (from lab to clinical research, and clinical research to practice, respectively). These barriers sub-divide the translational paradigm into three major components: 1) basic science; 2) clinical research; and 3) clinical practice / public health. If one reads the RFAs for federal funding opportunities that target the clinical and translational sciences as issued over the past several years, these sub-divisions have been reinforced, with requirements for such funding and associated projects to focus on the creation and validation of approaches to “breaking down the barriers” between such areas. The problem, however, is whether we are imposing a completely artificial model to conveniently break down the translational paradigm into “manageable” sub-components. For example, when defining new basic science protocols, it is rare for a translational researcher to not inform the design of his or her experiments in part based upon observations or previous work derived from clinical studies, practice, or the population sciences (for example, a scientist defining a protocol looking at certain bio-marker complexes in animal-based disease models might choose to focus on surrogate end-points or outcomes metrics that correspond to the human manifestation of the targeted disease). While a complete enumeration of the scenarios in which the sub-domains I have listed are more interrelated that the barriers between them would appear to suggest is too onerous to include in this post, I think the general idea is sufficiently clear. So what are the implications of these barriers? simply put, if we step back and look at the highly interrelated nature of the translational and clinical sciences, and our emphasis on “breaking down barriers”, one has to wonder whether we are attempting to fix a problem of our own creation, which is in effect a result of our own efforts to simplify a very complex system. If this is the case, perhaps we are missing the true opportunities to enhance translational science via the optimization of the interrelationships which already exist, rather than trying to catalyze new interrelationships that may not necessarily be needed. This might be an over simplified view of what is truly a complex system and problem, but certainly some food for thought when so many of us in the translational informatics domain are focused on effecting “transformative” change in our organizations - which begs the question, what are we trying to change exactly?

Translational researchers (e.g., basic scientists who move their discoveries into clinical trials, or clinical investigators who work to disseminate the results of their work as community-accepted guidelines or best practices) often look at the area of operations research as being a completely unrelated field to their research efforts. Instead, it is often seen as the purview of “business types”, such as organizational management or financial analysts. However, I would pose the question of whether this is in fact an appropriate division of focus, effort, and expertise. In this particular context, when I talk about operations research, I am referring to: “an interdisciplinary branch of applied mathematics which uses methods like mathematical modeling, statistics, and algorithms to arrive at optimal or good decisions in complex problems which are concerned with optimizing the maxima (profit, faster assembly line, greater crop yield, higher bandwidth, etc) or minimal (cost loss, lowering of risk, etc) of some objective function. The eventual intention behind using operations research is to elicit a best possible solution to a problem mathematically, which improves or optimizes the performance of the system.” (a definition as found on Wikipedia) When you connect the dots from last weeks post, there is a clear connection between analyzing and understanding workflow in the biomedical domain, and the optimization of that workflow, as can be achieved through the theories and methods associated with such operational research. While workflow analysis is a much more “fuzzy”, social-sciences anchored discipline, you can see how, given the preceding definition, operations research is where the scientific “rubber meets the road” in terms of acting upon the data generated by workflow modeling studies. Now - why is this important to Translational Research, and in particular informatics? Well, I would argue that our collective understanding of how new biomedical innovations or methods diffuse from basic science discovery to clinical research, and again from clinical research to clinical practice, is at the heart of enabling truly translational work. Basic scientists or clinical investigators, and their informatics collaborators, could generate the most amazing findings in the history of biology or medicine, but if we do not understand how to opitmally disseminate those findings, and maximize their uptake and application in the “real world” of clinical care or population-based sciences/services (such as public health), then those finding may end of becoming stale or forgotten in the academic literature or isolated islands of knowledge and practice. There are many reasons to believe that informatics, as a discipline, can serve as the mechanism by which operations research and translational science can be combined. As informaticians, we know how to instrument complex systems and collect high throughput data from a targeted environment, and furthermore, how to use that data in conjunction with other sources of knowledge (e.g., ontologies, terminologies, the published literature) and analytical pathways (e.g., hypothesis discovery, data mining, biostatistics, computational simulation). Such methods are critical to feeding the algorithms and models associated with operations research, such that we can understand system-based factors that affect knowledge or practice dissemination and uptake. However, to create such a research “pipeline” requires us (e.g., informaticians AND translational researchers) to recognize that to truly be successful in translational endeavors, we must recognize the need to engage in basic, clinical, AND operations research within the scope of a given project. Needless to say, this is a tall order, but one that I feel we need to embrace and act upon if we truly intend to change the face of biomedicine over the coming decades to focus upon data-driven, personalize health care delivery.

It seems that in the biomedical informatics community, we have a tendency to focus more on the technology that can be employed to improve a particular domain, rather than the “process” that we are trying to address. If we look to the publications and best practices that are available in the business and social science communities, there is a plethora of methods and approaches that can be used to support the study and optimization of such “processes” - often referred to as “workflow.” So the question is - why don’t we see this intersection being leveraged more frequently? Based upon some recent pilot projects that I have been involved with here at Ohio State University, I would suggest there are a couple of major reasons: The basic meaning of what constitutes “workflow” is not consistently understood or shared throughout the informatics domain. For those of us with more computationally-centric backgrounds, the term is used to refer to the orchestration of multiple computational processes, agents, or components. In contrast, for those of us with social sciences-centric backgrounds, the term is used to refer to a holistic understanding of the major players, artifacts, and activities that exist in a domain, and their relationships to one-another in a particular problem area. The tools available to support the observation, modeling, and analysis/optimization of workflow are diverse, and not always easy to evaluate or adopt. As with many “standards”, the real benefit of most workflow analysis approaches is that if you don’t like one of them, there are always numerous others to chose from. The literature is rich with descriptions of ethnographic studies, time-motion studies, key participant interviews, UML activity diagrams, “swim-lane” diagrams, and workflow simulation studies. Unfortunately, most of these reports assume a relative familiarity with their methodological approaches, making it difficult to quickly ascertain their applicability to ones own projects without extensive correlative research on those methods (if you can first find the correct resources from which that research can be extracted). One of the biggest issues with this lack of consistency and explanation is that methods and results from potentially complimentary studies are rarely compatible or interchangeable with each other. Imagine if everyone used a different way of calculating and presenting statistical significance in their publications (well, in all honesty, this is sort of the case, but I am not a statistician, so I will hold my comments on that topic for a later post)? That is sort of what reading the literature in the workflow analysis domain is like. Combinations of qualitative and quantitative approaches to workflow analysis seem to be rare, and in many cases, discouraged. The communities who frequently engage in qualitative workflow analyses often argue that quantitating such models will only lead to erroneous assumptions and conclusions. Similarly, the communities who frequently engage in quantitative workflow analyses seem to be solely focused on quantitating a domain, with little regard to the complex interplay between actors, artifacts, and activities. It seems to me that we need to find a middle-ground, such that these approaches can be combined and systematized. In all honesty, a fully qualitative workflow model is of little assistance on its own if we wish to instrument, optimized, or further evaluate a domain. Developing a better understanding of how qualitative workflow models can be linked in a efficacious manner to quantitative measurements of the domain under study should be a high priority goal for our field. And finally - how does this even relate to the domain of translational informatics? Well, from my perspective, it seems unlikely that we can be effective or successful in developing and deploying informatics platforms intended to address the needs of the translational research community if we do not first understand: 1) what the current workflow of our end users is; 2) where there are opportunities to optimize that workflow using informatics interventions; and 3) how we can design such interventions to achieve the targeted workflow optimization without disrupting other areas of the contributing workflow. Such an approach is consistent with about every well validated software engineering approach known - all of which aim to ensure that new software platforms are accepted and well utilized. The real challenge is not deciding to use such workflow or participant-centric approaches, but rather, the actual “nuts and bolts” of doing so in the context of complex biomedical workflows. The alternative approach is the “if we build it, they will come” model, which I think goes without saying and being less than ideal!

For those of you involved in the biomedical informatics community, you have probably noticed in the past few years that it is becoming increasingly common for both basic science and clinical investigators to invite informaticians to work with them on their projects. These requests can include assisting teams with the preparation of grant or contract proposals, to engaging in data management and analysis services that support the scientific aims of an ongoing project. One constant that I Ã¥have experienced in this regard, especially in the context of translational research projects (which tend to focus on linking bio-markers or biological models with phenotypic data in order to diagnose a disease or to understand disease progression / treatment outcome), is that the investigator community is often unsure what types of services or support they need from informaticians, other than that they need some sort of “informatics.” In a similar manner, I have found it to be a frequent occurrence that informaticians often have a hard time translating the novel theories, methods, and technologies they have developed or evaluated in order to address the specific problems of an applied research project. I often refer to this as the “then what” problem, as in, now that we have these technologies and methods, and a motivating use case, then what do we do? The reality is that this problem really points out a number of challenges that both the translational research and informatics communities are not doing a particularly good job of addressing, namely: Educating clinical and translational researchers as to the practical capabilities of modern informatics methods and techniques, Educating informaticians as to the realities of the “real translational research world”, Placing and emphasis on (and funding) applied informatics projects that can couple rigorous science, technology development, and real-world use cases, and Ensuring that informatics is understood by funding agencies, organizations, and funding agencies as being both an applied and basic science (usually in a simultaneous fashion), thus requiring it to be more than just a service, and more than just a theoretical exercise. Just some food for thought (which will probably be of particular interest to our colleagues, myself included, who are currently working on the ubiquitous CTSA application). I would be interested in anyones comments on how to address these areas (if you happen to be reading this).

The domain of translational research informatics is by most accounts in its early and formative stages - which presents a unique opportunity to academics and professional in the informatics community to define this timely and very critical field. However, that same opportunity also presents a challenge, due to the fact that a common perspective concerning what precisely constitutes the practice and definition of translational research informatics, or as I often refer to it in a simpler form, translational informatics, is generally lacking. Therefore, since this is the first of what I hope will be many posts to come on this new blog, entitled “Translational Research Informatics”, I thought I would take this opportunity to present one potential definition of the field that I, and a number of colleagues from AMIA (including my close collaborator, Dr. Peter Embi from the University of Cincinnati) have discussed over the past several months, and that we have posted to that ultimate of Internet knowledge repositories, Wikipedia (Translational Research Informatics). Our current working definition of translational informatics as found on Wikipedia, reads as follows: Translational Research Informatics (TRI) is the sub-domain of Biomedical informatics or Medical Informatics concerned with the application of informatics theory and methods to translational research. It overlaps considerably with the related rapidly developing domain of Clinical Research Informatics. Translational research as defined by the National Institutes of Health includes two areas of translation. One is the process of applying discoveries generated during research in the laboratory, and in preclinical studies, to the development of trials and studies in humans. The second area of translation concerns research aimed at enhancing the adoption of best practices in the community. Cost-effectiveness of prevention and treatment strategies is also an important part of translational science. Another way to look at the discipline of translational informatics is to visualize the major areas that must contribute to our collective ability to generate translational biomedical knowledge (which is concerned with understanding and acting upon the meaningful relationships between biological knowledge, clinical research/practice, and the population sciences). I have developed one such visualization, which I refer to as the “Translational Informatics Triad.” I have included the preceding definition as static-content on this blog to help guide readers who are new to the field. However, I fully anticipate that the definition will evolve over time, given our collective increased understanding of the field, as well as the specific comments I hope readers of this blog will contribute.