Eugene, Oregon
May 13-15, 1998
The Future Network: Transforming Learning and Scholarship
From Bench to Bedside and Beyond: Building a Regional Healthcare Network Test Bed
Sherrilynne Fuller
Acting Director, Informatics and
Associate Professor, Department of Medical Education
University of Washington
Today I want to share with you some views of what the problem is that is driving research and development in health care both on the technology and the computing sides. In addition, I will talk about some fundamental research issues and the research agenda in biomedicine.
I will use as my jumping-off point the regional telemedicine test bed that I am involved in developing at the University of Washington. And I want to characterize some of the issues in health care because at some point, like it or not, we all have encounters with the health care system. And so I think that probably my presentation will be in contrast to some of the others, but I think that may spawn a very interesting discussion.
To begin, I will discuss the health care environment in academic health sciences centers, which are truly undergoing revolutionary change which is not all positive. When you talk about the blueprint for a regional test bed, what are the component pieces? I will then say a few words about the President's Advisory Committee, and I will hopefully conclude all of this within 20 minutes.
As many of you have heard and read in the paper, we have a very dynamic health care environment: Increasing costs, decreasing reimbursements. Teaching hospitals by nature are much more costly because they are teaching hospitals. I will have trouble doing my talk on research and health care without talking about teaching; it is intimately interwoven. There is an increasing emphasis on training people to go out and do primary care and decreasing emphasis on specialists. However, primary care is not given in academic teaching hospitals very much. Most of our patients are the really sick ones. So, to do primary care training you need students out in the field learning from people who are seeing normal, typically healthy people. That is a change in how we educate and how we do research.
There are shifting referral patterns. Whereas patients used to come from and be referred to academic medical centers, this is changing and we are not getting as many patients as we once did. And this is true throughout the United States in academic medical centers. Finally, one of the hallmarks of the changed environment is increased interdisciplinary research. Health sciences research has always been very interdisciplinary, but it is even more so today.
What happened in the regional environment--and I am going to use the University of Washington as an example, but all academic medical centers have some degree of regionality: Either a state, an area, or throughout a number of states, as is the case at the University of Washington. The problem is getting patients to specialized care from remote places because of the cost of transporting patients; weather and geographic problems like mountains and snow in the Northwest; trying to deal with patient and family inconveniences; and, most importantly for the purposes of today's discussion, a very inadequate information infrastructure, a lack of connections, and a lack of systems resources.
The challenges for physicians, nurses, other health care providers are that they are seeing a lot more patients and they are more complex patients. Where they once referred a patient, now they have to care for the patient themselves. They are overwhelmed by the volume of research coming out. Every day there is something new in the paper, a new genetic test, a new piece of information; treat the disease this way, not that way. Patients are, in fact, keeping up better than physicians on their own personal illnesses, and that is a challenge for physicians. You go in and say, "I just read this article," but the physicians may not have read the article yet. And there is a lot of concern about the appropriateness of alternative therapies, in which many patients are very interested. So, again, it is a "keeping-up" kind of problem to deal with patients and consumers who know some and want to know more.
The researchers--and I'm including here both researchers on the clinical side who do patient research as well as basic research--are having trouble keeping up, too, with narrower and narrower specialization in terms of the research they do. They are competing for increasingly scarce dollars. They are trying to deal with large, interdisciplinary research projects handling hundreds of people, often working with researchers who are geographically distributed across the United States. And, again, effective access to knowledge resources is a core issue.
The WWAMI Program at the University of Washington, which is one of the earliest distance education experiments, began 30 years ago. WWAMI stands for the first initial of the participating states: Wyoming, Washington, Alaska, Montana, and Idaho. It is a very large area, about 28 percent of the land mass of the United States. There are points of presence of the University of Washington, either clinical or teaching or both, and research sites around the five states. All of these sites have to be connected in some way to us. That in and of itself is a very long story.
What we have been working on--and this is by the way an example of research and clinical teaching interrelationships--is a telemedicine test bed with a goal of evaluating a variety of issues related to delivery of care, supporting the teaching in a distributed way, and, most importantly, supporting research across a geographically distributed environment.
First, what is telemedicine? You probably have all heard the word. Very often people assume that it is simply using televideo consulting to link a specialist with a patient with a primary care provider, when the specialist is in Seattle and the patient is in Alaska with their primary care physician, for instance. But, in fact, it is much more than that. We define it both in terms of geographic separation as well as the use of any kind of communications and information technology to support either diagnosis, treatment, or monitoring of the disease or research. This definition is increasingly being accepted because the actual use of televideo consulting is probably appropriate in no more than 10 or 15 percent of the cases. The rest of the time the alternative is plain old telephone calls, which work just as well.
Images are critically important to telemedicine. Health care, medicine, and nursing is inherently an image-based practice. And without images--with only text--it is very difficult to do medical care. And therein lies the story of why high-speed, high-bandwidth type activity is vital to health care, and this area is where there is a lot going on at the present time.
Patient data: You want your medical record in the hands of the person who is caring for you when you have just had an accident, for example, but right now that is not an easy thing to make happen. We have developed an electronic medical record at the University of Washington, and we have embedded knowledge resources within that record. The point I want to make here is that it is important to have data about individuals in the context of knowledge in order to treat and care and diagnose those individuals' illnesses. The work that is going on in this area is very important both from a research point of view as well as for teaching and, of course, for patient care.
We are also working on making clinical mail more secure. Security is a very high concern in health care. However, we need to allow physicians who refer their patients to the university to keep track of the patient by seeing their record. Slide 13 is a test electronic medical record of a real patient. Information about the identity is disguised. This is the problem list--the problem this patient has is retrolental fibroplasia. The point about this is that there is information behind each of these clicks: Radiology reports, labs, and so forth. But the important point is that the icon at the front is the embedded knowledge resource I mentioned. In this case, clicking on the "i" icon automatically searches the system at the National Library of Medicine and pulls up all relevant citations about this particular condition. Another click will actually order the document from the library and bring it into your hands. This is not a prototype system--it is in action throughout the University of Washington medical centers.
In the medications area we also have a knowledge system (see slide 15) that has been developed to support physicians who are writing prescriptions and needing other kinds of information, for instance, on adverse effects. Again, clicking on the "i" icon brings up a set of resource databases, including basic information about the drug toxicity and so forth.
So, the research into integrating knowledge resources into teaching, care, and basic research is very much a goal of what we are about and what is going on nationally and internationally in terms of health care. This is blurring the distinction between the library and clinical information systems, and it is providing librarians with some very interesting opportunities for involvement.
In our case, we are the regional library for the Pacific Northwest under contract with the National Library of Medicine. Our region includes the WWAMI except Wyoming but including Oregon; we are working on digital library initiatives in a variety of areas. The goal is bringing knowledge to the point where there are answers to questions, not pointers to answers.
One of the knowledge resource approaches that is supporting research, teaching, and care at the university is a web approach to information for the health sciences. This is a joint project of faculty and librarians. The tool kits are available for a variety of roles that individuals have either as students, grant seekers, instructors, or researchers. And the researcher's tool kit provides pointers that provide access to actual knowledge resources across the institution.
The molecular biology and genetics area is one of the basic research areas in health care that is burgeoning, creating many, many different tools and resources and databases. How we array these, how we provide access to these, and what will happen in the context of the discussion about copyright is anyone's guess. But at this point there is much more capability available to support researchers than the researchers themselves can possibly keep up with. That is one of the purposes of arraying the knowledge in the way that we were developing.
Finally, just to say a word about the subject of the advisory committee. The title is very long, but in a nutshell it is an advisory group to President Clinton: The Advisory Committee on High Performance Computing and Communications, Internet Technology, and Next Generation Internet. Our focus and the goal in the appointment letter says that we are to guide the federal investment and basic research on computing information and communications. So we have a number of themes. The basic themes of our deliberations include a need to enhance and support the basic research base, and that we are not funding enough research that is looking far ahead--out ten years. We need to worry about the fact that if there is a shortfall in the workforce, companies cannot hire people, and universities lose students to industry before they ever get to graduate school. There is a need for strategic vision for the long term.
We have also talked a lot about the need for more interagency collaboration in the federal government between the lead agencies in information technology computing: The Department of Energy, the National Institutes of Health, NASA, the Department of Defense, and so forth. This interagency collaboration amalgamation is something that could be improved upon--I am speaking for myself as an individual on the committee; I do not speak for the committee.
Let me mention some of the technical hurdles we now face--and this is in terms of health care as well as regarding computing communications in general. We need on-demand, high-bandwidth, high-speed, robust kinds of computing. We need to know that the network is available when we need it, 24 hours a day, seven days a week. There are problems with feedback latency over distance, being able to have true synchronous communication, and links for distributed databases. Your patient record may be in many places, in clinics and several hospitals, and if you are in Europe when you get sick, you want those pieces of your record to come together in some way.
Then, one of the areas that I am particularly interested in but have no expertise about whatsoever is how to represent other senses. Physicians need more than visual sight of the patient to assess condition. They say they need to smell it, they need to touch it. And, certainly for virtual surgery, they need feedback to be able to sense pressure. Other technical hurdles are content-based imagery from 2-D to 3-D databases; security; integration of software, so that you do not need 20 devices to do a single telemedicine consult; and better devices.
In conclusion, there are many challenges facing us as academic health sciences centers and researchers. We need a state of the art national information infrastructure. The problem is that, so far, health care has really not risen to the occasion and been a party to the deliberations. There are only two people from the health care arena who are on the President's committee, myself and Ted Shortliffe at Stanford University, a physician. And the National Institutes of Health is, in fact, not represented in that important area, although it is beginning to become more involved.
I think the goal for all of us is just-in-time information at the point of need. Here are some web addresses for further information about healthcare delivery and research in the networked environment:
Thank you very much.
Accompanying Powerpoint presentation
Share
Twitter
Facebook
YouTube
Flickr
Assessment Blog
Policy Blog
E-mail
RSS