In late October, the University of Chicago established a new degree program, creating a master of science degree in biomedical informatics. The program is being offered by the University of Chicago Graham School of Continuing Liberal and Professional Studies, and has brought together faculty members from genomic research, translational medicine, and computation, to offer the Master of Science Degree in Biomedical Informatics (MScBMI). The program combines “in-person instruction and industry-based capstone projects, made possible through relationships with healthcare organizations like Blue Cross and Blue Shield of Illinois and nonprofit institutions such as NORC at the University of Chicago,” according to an Oct. 28 press release from the university.
As the university’s press release notes, “The MScBMI is designed for working adults. The part-time program offers an individualized approach to each student providing direct access to discussion-based classroom instruction and team-base experiences,” and has two faculty directors, David McClintock, M.D., an assistant professor of pathology and medical director of pathology informatics; and Samuel Volchenboum, M.D., Ph.D., associate professor of pediatrics and director of the Center for Research Informatics at U of C.
Here is a link to the program’s core curriculum. As the program’s website notes, “Our curriculum has undergone scrutiny by faculty and industry partners to ensure its relevance and applicability to the current workforce needs in biomedical fields. The goal of the program is for students to learn and master the following: informatics methodology, applying tools and techniques to both research and applied problems in biomedical settings; effective communication with diverse professional audiences regarding informatics issues and solutions; management of biomedical informatics projects; understanding of the ethical, privacy, and data security issues in the field.”
Shortly after the inception of the program, Dr. Volchenboum spoke with HCI Editor-in-Chief Mark Hagland regarding the program specifically and the development of degree programs in medical informatics more broadly. Below are excerpts from that interview.
Congratulations on the launch of this new program. How long have you been involved in its development?
I’ve been involved in the programming of it for several years. And education is part of the mission of the group I run here at the hospital.
Where are things in terms of process right now?
People are being admitted to the program now; the first class will start in spring, in March.
How did you and your colleagues end up developing the concept for the program, and the program itself?
I’m a pediatric oncologist by training. I do a couple of weeks on service a year, and still do a clinic once in a while. I did an informatics fellowship at Boston Children’s Hospital, part of the Harvard Medical School/MIT program. So that was a tremendous opportunity for me to go back to school, and to advance in informatics, which I was always interested in. I did a pediatrics residency, followed by an oncology fellowship, and then an informatics fellowship. Then I took on a faculty position here at U of C, running a proteomics research lab.
Then I was presented in 2012 with an opportunity to help lead the Center for Research Informatics. Since 2012, I’ve been the center’s director, and we’ve grown into a group of 40 people that provides the informatics support for the Division of the Biological Sciences, which includes the research enterprise and the medical school.
The 40 people you’ve mentioned make up the research informatics group, then?
Yes, they’re all here to support research informatics. We have nine Ph.D bioniformaticians who do high-throughput genomics, for example. We do support research for the labs, as well as performing our own original research. Among other things, we are analyzing high-throughput, next-generation genomic sequencing data.
What’s the context for the evolution of formal education in bioinformatics?
When people talk about bioinformatics, they are normally talking about genomics and other kinds of computational analyses of large sets of data. A bioinformatician generally has training in the kinds of statistics and other kinds of math to do genomic sequencing or mapping. Bioinformatics, however, also includes medical informatics, which is related to patient care. While I have a foot strongly placed in both areas, most of our Graham School program is concentrated on biomedical informatics.
Historically—and even currently—biomedical informatics is something that most people have learned while doing it, correct? And this is not something that most CIOs or CMIOs in patient care organizations have much practical grounding in, right?
Yes, and per CIOs and CMIOs, we have a whole generation of physicians who are being asked to do jobs in the medical center for which they’re not sufficiently trained. It’s providing a lot of opportunities, but it’s not always ideal to learn on the job. It’s good and bad: you learn by doing projects, but at the same time, you can be prone to making mistakes.
But obviously, it is your belief that more formal biomedical informatics training is called for?
Yes. Institutions are now accepting that data are some of their most valuable assets, and the use and misuse and protection of data can make or break an academic medical center. So institutions are working hard to put together programs—governance programs, analytics programs, data warehousing—that are needed. And these will be needed in ways that doesn’t exist in most places. For example, we’re building out our governance structure to make sure the data being used are the best data possible. I’m leading those efforts, because I have a background in understanding what data governance is. The other physicians involved are very smart and experienced, but don do not have the background or formal training in biomedical informatics. Having this kind of formal training can really help a lot.
Another good example is the way people see research data: if you’re collecting data for a research study and have no background in how to collect data and store it, you can find out at the end that you should have done things differently.
Is this the first program of its kind in the country?
There are programs already established elsewhere. There’s a page on the AMIA [American Medical Informatics Association] website that shows the accredited programs, and we’re aggressively going to pursue accreditation. But there are a number of programs nationwide.
What core elements are common to these programs?
Most programs will focus on the core curriculum, as defined by AMIA, including information on electronic medical records (EMRs), and how the EMR requires integration, interoperability, and governance. Also important are issues of meaningful use and other governmental regulations. And someone headed towards a CMIO career will need a healthy dose of understanding how those systems work. The backgrounds of the students are diverse. You might have someone coming in who works for Epic but has very little clinical experience—so trying to teach them what happens on rounds might not be that easy. Likewise, someone coming from a clinical background—a cardiologists, for instance—may have had little or no information technology background. So we’re trying to “level the playing field” by offering the right kinds of courses and “boot camps.” In addition, we’re requiring a formal computer programming course to give our students a solid background in coding.
So a part of this involves taking clinicians and teaching them some core informatics skills, correct?
We are enrolling both clinicians and those with backgrounds in statistics and computer programming. One needs to be careful about the how clinical scenarios are presented, so that the technical students can follow along. But by exposing both groups to technology and clinical practice, we’ll create a next generation of people skilled in both things.
Hybridizing people, then?
Yes, and some organizations have created two tracks; we’re trying to get everybody on the same page to start with.
What developments on both the research and patient care sides do you want to help people prepare for?
Great question. The collection of the data is really becoming commoditized. There’s more data than ever before. The key in the next ten years will be understanding how to use and apply data. You can now have a genomic panel test to see which drugs are likely to be more effective for a patient or you might be able to learn about how quickly a patient may metabolize a drug. So, that’s one extremely important, active area, and I’d be surprised if clinicians aren’t soon offering genomic panel testing done in conjunction with the delivery of your routine medical care. The real explosion right now is in cancer genomics. It’s very common to have a tumor biopsy sent for panel testing. The thing is, we’re great at sequencing tumors and quite good at the analysis of those data (lining up the sequence data and understanding where the variations are), but you can get those results, and then a pathologist needs to come up with a set of recommendations that will help the patient.
There is so much rich information now, and trying to link these variations in the genome to prognosis and patient care is key. For example, a little boy I take care of has a desmoid tumor, and we got a lot of information back from that sequencing. The choice was difficult, then, because there were plenty of drugs you could apply to the situation, but the actual data about which drugs will be effective is still missing. Additionally, we’re using sequencing information to help learn what drives the development of tumors. So until we get enough evidence on which drugs would help, it’s still very empiric on how to personalize the treatment. Our group just developed a suite of software tools to aid molecular pathologists in generating their findings and reports. This platform allows our pathologists to record their findings and then propagate those data forward to subsequent reports, where appropriate. This will facilitate another level of study, whereby we can learn about the effects of mutations over entire populations.
Have you developed the full curriculum?
Yes, we’ve fleshed out the core curriculum and the courses, and we’re securing the instructors. Here’s a link. We’re also trying to build some courses that leverage specific strengths of the University of Chicago, such as ethics and legal issues in biomedical informatics.
What should our audience understand about where all of this is headed?
I think they should understand that there is a gap between the amount of data we have and how to use those data. This gap can be filled through training programs like ours. At U of C, we’re trying to make it very feasible for students to complete their studies while they are clinical fellows or continuing their work on other projects. I think that’s going to attract students who want to go pick up on this vast array of tools. The current paradigm is running out of steam. Ten years ago, you would find some clinician who had taken a little computer science class on the side, and you would say, hey, you’re going to become an informaticist. But the world is becoming more and more complex, and someone could apply these tools incorrectly and worsen a situation. These kinds of programs will help alleviate these risks – especially for those destined for leadership positions, such as CMIO and CIOs.