Informatics: Balancing Application and Theory

By M.L. Baker  |  Posted 2004-10-08

Informatics: Balancing Application and Theory

How does a field define its essence, when being interdisciplinary is essential? Could doing so make it less powerful?

Those are central questions for informatics, the art of using information and information technology to uncover relationships and to optimize decisions. "Its not really just about technology; its about how people use information and how they use technology," said William Hersh, department chair of medical informatics and clinical epidemiology at Oregon Health and Science University.

Business and legal informatics exist, but as countries push for IT solutions to make patients safer and care cheaper, the prominence of health informatics is growing, as are arguments about what exactly it is.

"Were still trying to hash out what those [core] competencies are. Most cardiologists could tell you what you have to know to be a cardiologist. In informatics, there are always quibbles," said John Holmes, chairman of the education committee at the American Medical Informatics Association and professor of medical informatics at Pennsylvania Universitys School of Medicine.

That doesnt keep Holmes from listing several things every informatician should know: Besides domain knowledge, "people need to know about information systems theory [and] programming skills. You should be able to write a small program in a reasonably modern language, know the key areas of research, clinical vocabularies, decision-support systems, have grounding in evaluative sciences."

Another problem is that advances in computer and medical science can quickly drive technologies and skills into obscurity. (Alan Turing didnt use C++ for his breakthroughs.)

Perhaps most controversially, the optimal blend of informatics and "domain knowlege" is ambiguous; among different universities, there is a continuum of views about the degree to which general principles of informatics can be abstracted and the degree to which such principles are inherently linked with specific disciplines and applications.

Holmes said the university issuing the degree would influence students future career options. "Barring some sort of standardization process, where you went to is pretty important, because we know whos gone there." Of course, students tend to select schools based on the training they want, but at least three of the people interviewed thought students are not particularly savvy about these distinctions.

However, in masters and doctoral programs that require independent research, students can make a theoretical program more applied and vice versa, according to Russ Cucina. After completing his medical residency, Cucina entered Stanford Universitys Biomedical Informatics Program. That has a reputation for more theoretical training, but he sought out more-applied projects.

Cucina is now a professor at the University of California, San Francisco, but just over half of his responsibilities include clinical work and helping the university hospital transfer from one vendor of clinical software to another. These will keep him from becoming "too distant" to be able to bridge clinical and IT work.

Cucina said part of the appeal of Stanfords program was that he had already established useful relationships during his medical residency there.

For many students, logistics weighs as heavily as a particular programs theoretical or applied bent.

Michael Mundorff, a clinical programs analyst at Intermountain Health Care, said he has been considering medical informatics programs since working on a biosurveillance project where he saw such expertise prove particularly useful. "Having been employed in various capacities in the health care field for more than 20 years, I would prefer an applied-focus program to a theoretical one," he said. He referred fondly to the very well-regarded program at the Department of Medical Informatics at the University of Utah. But hes also looking at the distance-learning options at other schools with the NLMs (National Library of Medicines) blessing. "Im not sure that attending campus-based classes would be compatible with my work schedule."

Next page: Degree "just one piece of the puzzle."

Page 2

Laura Gunn, recruiting manager at CapGemini Health, said a persons degree is just "one piece of the puzzle" when considering job applicants. "We are primarily interested in those programs that enable their students to quickly and effectively contribute as consultants and produce results, whether the program is theoretically oriented or transactionally oriented," she said, adding that candidates also need real-world experience, and that the opportunities students created for themselves in educational programs could factor heavily.

Russ Altman, director of the Biomedical Informatics Training Program at Stanford University, said a theoretical underpinning provides the most powerful options. "The most important thing is to understand the core methods for manipulating information. Then you can apply it to different situations," he said.

"The translation of theory into practice is a difficult problem. Some argue that this is best left to industry, where they have resources to do it right, and the reward system is more clear for a well-engineered practical system. I tend to agree," he said, adding that some of his colleagues at Stanford would not.

Students choice of school impacts their future careers. "Individuals take on responsibilities largely influenced by their training," said Joseph Hales, director of information systems/medical informatics and a professor of medical informatics at the University of Utah, hastening to add that the trends were small. "Those with an applied background seem to focus more on running systems, using EMRs [emergency medical records] as a tool to improve safety or process, building tools that make systems more useful. Those with a more theoretical background often take a more academic approach, even if in an operational or commercial setting. They are designers, working to enhance fundamental understanding or demonstrate the utility of advanced features of EMRs.

"Programs that are more theoretical have historically produced more informaticists who stay closer to academics, either in a faculty position or at academic medical centers, while programs characterized as more applied have historically produced more informaticists that have moved on to industry." But these distinctions are fluid, he said, since applied informatics can quickly veer into the theoretical when trying to solve difficult problems.

COMING SOON: Bioinformatics and medical informatics. One discipline or two?

Program names are emblematic: "Health informatics" is considered a fairly broad. Some feel that "medical informatics" excludes non-physicians like nurses and that "bioinformatics" is an entirely different discipline. The 18 programs supported by the NLM use at least eight different descriptors; "biomedical" and "medical" are the most common. The word "and" as in "biomedical and health informatics" or "biology and medicine" or "computation and informatics" recurs.

Check out eWEEK.coms Health Care Center for the latest news, views and analysis of technologys impact on health care.

Rocket Fuel