Content: In reviewing the lectures of 8/31/09 and 9/2/2009, it is clear that we are at a critical time in a pivotal department. After years of seeking an identity for the field of biomedical informatics, federal appropriation through the American Recovery and Reinvestment Act has established the credibility and importance of this field. The Greenes lecture, by one of the founders of the field, in combination with his recent article (Greenes et al, Academic Medicine, 2009) emphasize that the field incorporates a broad variety of foci from molecular to population levels, that research that is interdisciplinary is important, but that the field is applied and a combination of “IT systems, human beings, and organizations aimed at achieving a particular purpose.” As such, this latter view differs from the more “computational” view of biomedical informatics that is advanced by Bernstam et al, “Synergies….” (Academic Medicine, 2009). “BMI utilizes IT and computer sciences as tools and applying such capabilities, but it is driven by a deep understanding of the science and practice, problems, interactions, culture, and milieu of biomedicine and health.” Greenes et al Academic Medicine, 2009). Biomedical informatics is distinct from health care informatics and although computation is important, but diminish the scope of BMI (Greenes, 2009). He further points out that clinical data repositories and clinical data warehouses, the latter for research application are important and that BMI professionals are uniquely suited to work in both areas. The broader approach by Greenes incorporates the narrower focus of Bernstam et al as to the utility and components of IT as part of research efforts involving the NIH funded Clinical and Translational Science Awards that are trying to bring bench to bedside. Bernstam et al define the groups as Operational IT, a more central institutional group, as applied computer scientists and local research IT groups that fuel, maintain, and innovate for specific research applications. PhD computer scientists do research on questions of technology, such as the physical handling of data, given its enormity. Biomedical informaticians are focused on research and supporting the goals of the CTSA at the University centers. They are concerned with information and knowledge management and developing new ways of managing information (Bernstam et al, 2009). This present a more limited focus for BMI. However, close cooperation is needed between operation IT, research I and biomedical informatics. A clear administrative structure for each component and a clear role within the specific institution is needed (Bernstam et al, 2009). The issue of training BMI individuals in programming is brought up and I would suggest that this is critical not only for supervision and interaction, but also for a broad understanding of the field.


One aspect of BMI is the generation of research which can lead to applied interdisciplinary solutions. This can involve models of reality, manipulations of reality that include simulations, and empirical studies. Models of reality may involve different types of decision making/problem solving processes. BMI involves all and each of these components.

As Greenes defines these essential points, Dr Fridsma, who has developed the software for generalizing clinical trials, tackles the difficult issues of ontologies. Are these truths or not truths? This depends on the issue and the perspective. Controlled vocabularies are built into taxonomies and then made more explicit in ontologies. If biomedical informatics principles are to be applied, then generalizable languages and definitions need to be established and used uniformly. SNOMED represents a more generally accepted language, in part due to its post coordinated aspect, as opposed to ICD9 and DRG coding. I use ICD9 codes so that I can be compensated. Hospitals use DRGs. These languages are not necessarily the same and certainly can not be used together in research applications. Fridsma also distinguishes database schemas, focused on data, metadata, and data dictionaries versus knowledge based ontologies. It is clear that we need more references on these areas covered in this lecture. The UMLS provides a way to merge and construct specific vocabularies and has tools to do this as well as services. Until there are more uniform languages and vocabularies for medical informatics, it will be difficult to implement successful programs for improvement of health care through informatics, despite the large new financial allocations. Any thoughts? That some of these lectures have been recorded by Mithra and made available to us is very useful. I hope that this can continue throughout the course.



Stuart