Information and Communications
Don E. Detmer, M.D.
Senior Vice President, Louis Nurancy Professor of Health Sciences Policy, and Professor of Surgery, University of Virginia
I greatly appreciate this opportunity to share my sense of what lies ahead, including a few ways in which the Institute of Medicine may help shape the information age. My discussion is about looking forward to 2020, which is a presumptuous task. What you might get instead is a "my-opic" view of 2020. At any rate, we can say that there is something quite exciting coming.
Lewis Thomas would remind us that our biology is based upon a carbon structure. Now we are entering an era in which we will fuse carbon-based intelligence with capabilities that are silicon based. We are crossing the threshold into a new chapter of human existence. What follows are my speculations on this inspiring fusion.
Vannevar Bush stated some years ago that the world had arrived at an age of cheap, complex devices and great reliability, and something was bound to come of it. The status was no longer quo. As Bela Bartok put it, "What is new and significant must always be connected with old roots, the truly vital roots that are chosen with great care from the ones that merely survive."
"Bite by bite and bit by bit," as Robert Kahn said, "eventually adds up to quite a bit." Today, computers can store and retrieve a thousand-trillion bits of data, and the expectation is that there will be performance at the 1 million-trillion level by the end of this decade. Petabytes and etabytes are quite a bit.
The big picture, as I see it, is in terms of both information and communication, comprising the medium—the telecommunications infrastructure, as well as the message—managed care capitation, biomedical research innovation, public accountability, and the topic that I am principally addressing, how all of this fits together.
The capability will become powerful enough for some of these distinctions to blur at the margin. So I want to mention a few high points related to the information infrastructure itself, the medium, and then focus on a few points related to the messages before offering some philosophical reflections on computer ethics.
The major message for health professionals using a telematic medium will relate to managed care (capitation), biomedical research and innovation, public accountability, and the technology itself, including the computer-based patient record as envisioned in the 1991 IOM report, The Computer-Based Patient Record: An Essential Technology for Health Care.
The infrastructure is changing rapidly and, at the same time, not as quickly as people perhaps would like to think. We are seeing more personal computers with mainframe power, scalable parallel high-performance computing systems capable of monitoring and maintaining a superpower's nuclear force without future actual testing, and the emergence of a global neural network connecting the U.S. national information infrastructure to the G7 (Group of Seven) global information infrastructure.
More than its share of leadership for the global information infrastructure is coming from the United States, stimulated by the development of the U.S. High-Performance Computing and Communications Program (HCCP) enacted in 1991. This act, championed in 1989 by Vice President Gore, then Senator Gore, has given focus and integration to the computing initiative of 10 separate governmental agencies. As most of you know, IOM Council member Donald Lindberg was the founding director of the HCCP and as head of the National Coordinating Office gave health applications the visibility and the attention that they certainly deserve.
Four Transformational Forces
1. Managed care or capitation systems that place providers at risk.
2. Biomedical research and innovation.
3. Public accountability.
4. Telemedicine and the computer-based patient record.
If we look at the global situation today, we see a patchwork of incompatible communication networks marked by high-cost, low-quality services and very limited interoperability between systems. However, that infrastructure is changing with the emergence of broadband transmission.
Broadband transmission—the cable, wire, and wireless transmission of multiple frequencies of more than 2 million bits per second—is the cement that will bond communications and computing. The major issues about the technology itself are still not settled. For instance, will the communication mode be in bits routed through asynchronous transfer mode (ATM) switches or in waves from variable-wavelength lasers. According to some experts, if the bit mode prevails, fixed-length, 53-byte cells will be routed by way of ATM switches, which will combine the best of circuit and packet switching with efficient bandwidth guarantee if the standards are in place. The estimated time of delivery is 2000 or 2010.
Alternatively, optical ether transmission in wavelengths tuned to varying wavelengths by lasers would travel instantaneously by way of electronic switching equipment, with the bit technology totally eliminated. The estimated time of delivery for this is 2010.
When will multimedia really be here? Without demand, truly global broadband networks could be decades away; it depends upon consumer demand for services. Short of massive government subsidies, American-led entertainment and recreation are likely to lead to its development, with business applications and health care following in succeeding waves. The truly ubiquitous broadband networks across the United States could even take 30 years.
So, despite the challenge of creating a global network, we are seeing principles for a global information infrastructure emerge, as well as global organizations such as the International Association for the Advancement of Health Information Technology, which was initiated in Geneva this fall. Overcoming the political, cultural, and financial barriers is part of the challenge, but leadership remains the most crucial aspect. We need more leaders such as Don Lindberg, Jan H. Van Bemmel in the Netherlands, and Marion Ball, who recently completed an outstanding term as president of the International Medical Informatics Association.
The principles include interoperable systems and applications, privacy and data security, protection for intellectual property rights, universal access
to networks, and the encouragement of research and development. Despite what is yet to be done, much is happening. For example, the Internet is expanding rapidly. In March of this year, the High-Performance Computing Center reported that there were 27,000 networks, representing a 350 percent increase in the past 12 months alone.
Forty countries are now connected by way of the National Science Foundation's International Connection Management Program. Estimates vary widely from 10 million to 24 million people who are now currently connecting with the West, but no one denies that the number continues to grow.
Telematics is a recent term used to reflect the impact of the coming broadband capability. Whereas the video component today is 10–15 frames per second, camera and voice require 90 megabits a second for a normal appearance.
One way to define health telematics is that it encompasses both telemedicine and medical informatics or health informatics. Telemedicine and its evaluation are topics of a current IOM study chaired by John Ball.
In addition to teleconferencing for educational purposes, telemedicine also helps transmit computer-based multimedia patient records for consultation over large distances. Early experience suggests that substantial dollar savings are realized by better triage decisions regarding transfer of patients to higher level care facilities.
Virtual surgery is also in the wings. In Finland a few weeks ago, a surgeon from France reported at a meeting I attended that a pig in San Francisco had its gallbladder removed by a surgical team working in Texas and Oregon. I have been unable to confirm that story, but whether it is true or not, virtual surgery is on its way. Beyond treating acute appendectomies on submarines, it is not quite clear what the implications of this might be.
For those of us who live in states with substantial rural populations, telemedicine offers a real potential to get health care services to isolated communities and into people's homes generally, particularly for the elderly.
Medical informatics is the use of computers and information sciences to facilitate clinical practice, medical research, and teaching. Examples of applications include knowledge bases, calculations, data retrieval and display, decision support and evaluation, synthesis, and analysis.
The growth in the knowledge base of medicine alone is prodigious. Reports by John M. Williamson and others have shown that only about 10 percent of primary care physicians keep up with the medical literature. I cannot find similar studies on other types of physicians, but it is clear that only with information and communication technology can both generalists and specialists, as well as all members of the health team, hope to stay current with the exponential growth of the knowledge base.
Further, it is increasingly impossible to keep up with the paper and the digital information glut, especially as we live in this messy "metastate" with both paper and digital systems. Our drive to become fully digital is now in-
creasing in fervor as we struggle to reduce the deluge of overlapping and duplicated information.
Herbert Simon, the Nobel Laureate, captured this nicely:
What information consumes is rather obvious. It consumes the attention of its recipients. Hence, a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.
Where do we place our attention so that these information and communication technologies can best improve human health? This is going to be the balancing act for the next 25 years.
Clearly, we need to relate telematic technology to the messages of our environment: managed care, capitated care, biomedical research and innovation, as well as public accountability. Yet, more specifically, how do we do this—particularly since the complexity of our world is such that we will not control it?
For example, all of us are managing, or trying to manage, managed care, but it is obvious that we are dealing with forces we do not ultimately control. As Woody Allen says, "The lion and the lamb will lie down together, but the lamb won't get much sleep."
So I want also to discuss briefly the shadow side of this. Will the medical marketplace focus on price or value, on profit, or on access to needed services? Will telematics be built to serve the economically advantaged part of the world or everyone? Will its software programs achieve the stated goals or simply represent "vaporware," that is, software whose performance does not come close to its advertised value? Will our standards of accountability in medicine be our clinical impressions, or will they be based on hard evidence? Finally, will our patterns of interaction reflect continued paternalism or maternalism, or will we use these information communication technologies to develop real partnerships with our patients and society?
I would posit that between now and 2020 we have a grand challenge with two component parts. The first is to mature the health evaluation sciences: biomedical science, biotechnical sciences, ethics, biostatistics, clinical epidemiology, health services research, and telematics. The second part of this grand challenge is to integrate the health evaluation sciences with the evolving biological and the psychosocial science base.
I think we can attend to that challenge only by capitalizing upon information and communication technology. This will require us to develop and manage knowledge bases, to manage value, to ensure accountability through evidence-based medicine, to integrate patients' desires into our practices and public reports through the development of computer-based records and computer-based record systems, and finally, to formulate methods and policies to critique our goals and values.
Any sensible development of knowledge bases must have robust vocabulary servers with a unified medical language. A multinational effort mounted
Medicine in 2020: Grand Challenges
Advance the health evaluation sciences, including:
• Biomedical/biotechnical ethics
• Clinical epidemiology
• Health services research
• Telematics (telemedicine and health informatics)
Integrate the health evaluation sciences and the biopsychosocial science base
at this time would avoid a great deal of effort to pull divergent approaches together in the future. A major effort begun now could essentially be complete in five years, including setting up a process to maintain it into the future. The IOM could play a major role in this important global development.
Knowledge bases at various levels and for a variety of purposes already exist. These include some, however, totally new and unique data sets. For example, with emerging on-line communication, observational data shared among patients themselves are developing real credibility, particularly among the patients. What are we as professionals and the patients as patients to make of these databases? Can we examine these data and reshape them into knowledge bases that specialty societies take on as part of their mission? Clearly, the task of maintaining knowledge bases is likely to be even more daunting than their initial development.
As the excellent IOM definition of quality of care illustrates,1 if we are to manage to value instead of price, the clinical knowledge bases will never be completely static. Current professional knowledge is certain to be a swiftly moving target because of the contributions of biomedical research and innovation. Maintaining and evaluating our knowledge bases will require constant attention. Specialty organizations again should serve society by performing this admirable function. In addition, as the World Wide Web becomes filled with all manner of health-related data, whose role will it be to sort out the signal-to-noise ratio?
With respect to the delivery of health care as we know it, greater public accountability and involvement are needed in the care of individual patients, as well as in reporting the results of the care of groups of patients.
Report cards on organizational performance, such as those being issued regularly in Cleveland, come to mind. Currently, these require a great deal of effort involving paper systems. The computer-based approaches that follow this will be more efficient and complete.
Only with robust information systems can we get more science into our procedure-based clinical work. Evidence-based medicine of this type has had three principal reports: Archie Cochran estimated in 1976 that less than 10 percent of medicine is currently grounded solidly in reliable data; the Office of Technology Assessment (OTA) in 1983 estimated 10–20 percent; and NIH in 1990 estimated 21 percent. Whatever the exact situation, the situation is improving but still has a way to go.
The greatest impact of information technology on practice should come from the momentum developing behind evidence-based practice. This will involve automation of a number of care processes, and only time will tell how quickly these will develop. With the proper set of incentives and data systems, the changes could be surprising.
The possibilities include technology assessment, management of diffusion, clinical guidelines, and the debridement of useless procedures from studying what comes out of computer-based records and outcome analyses. The proposed IOM study on assuring clinical proficiency is very timely in this regard.
The Institute's 1991 report on computer-based patient records laid out a vision that included a primary recommendation that such records become the health records of choice. That report, with its 12 attributes for computer-based records and their systems, has generated a great deal of momentum, and the report of another IOM committee chaired by Roger J. Bulger, Health Data in the Information Age: Use, Disclosure, and Privacy, published in 1994 maintains the momentum. While progress continues, it is likely to take until 2020 for widely available computer-based records to meet all of the criteria developed by the IOM committee.
One of the recipients of the 1995 Nicholas Davies Award for Excellence awarded by the Computer-Based Patient Record Institute was the University of Utah group at the Latter-Day Saints (LDS) Hospital in Salt Lake City. Reed Gardner spoke directly to what his group has learned from two decades of effort in building computer-based patient record systems. These were his messages: First, improve the process and remind yourself constantly that the patient is central—not the doctor, not the nurse, not the administrator, but the patient. Build systems to prevent problems, not to detect problems that you then have to figure out how to prevent later. Focus on the system, not on the individual. The Utah group also concluded that variation in clinical practices is endemic and that quality can be improved constantly.
Evidence would suggest that impressive gains can be achieved with fairly simple systems that let you practice good medicine but remind and
support you until the task gets done. At the individual clinical level, this will also involve development of computer-based decision support systems that integrate the individual patient's utilities into the patient's care plan. This is real accountability at the point of delivery of care.
William Knaus and I have recently discussed how this might be done to improve the care and decision support for dying patients, a need highlighted by the Robert Wood Johnson Foundation-funded study that he codirected, which was recently reported in the Journal of the American Medical Association.2
Now, let's return to the big picture, keeping in mind Bela Bartok's comment.
I have hinted about a "pro-fusion" of applications and considerations; alas, there will be some "con-fusion" as well. What follows are my speculations. First, I think that society will reformulate around new functional aggregates of geographically dispersed individuals. I am talking about virtual reality becoming actual reality for these subgroups through use of the World Wide Web.
A new, powerful form of ethnicity will develop around telematic-based work and function. This will have its geographic movement component as well. Keep in mind that each day, 1 million people cross national boundaries through air travel. We are just beginning to see elements of this, but I do not think it is yet truly developed. This development will eventually challenge some aspects of nationhood and of cultural and religious heritage.
Cultural concepts will begin to be redefined: for example, democracy, health, and autonomy. A new reality and a redefinition of democracy may emerge that will be a more true democracy rather than a representative one since communication technologies allow for more direct governance, an impossibility without this technology.
Health itself may become redefined to include not only one's unassisted and unaccompanied physical and mental well-being but also a well-being that is intrinsically supported by a number of artificial computer-based capabilities. The inanimate becomes incorporated as essentially being or becoming partly animate. For example, artificial hips, pacemakers, and even imbedded intelligence may emerge. Is a person "healthy" even if he or she is healthy only with a host of new body parts?
Then there is autonomy. When a decision support system mandates that a doctor take action in a tense clinical situation, yet does not reveal how the system arrived at its directive, who is really taking the action, the computer or the physician? Where is the autonomy and where is the responsibility?
How much of our sense of the word "knowledge" will become based not upon the reliability of something's appearance, which we have generally held as a criterion of knowledge, but instead upon meeting a more rigorous test of not only reliability but validity?
Consider the blurring of the margin of animate and inanimate, of perception and reality. If you are a surgeon operating in a virtual telesurgical environment, what is real—the image you are working with on the screen and what you assume is happening on the other end, or what is actually happening on the other end? Is perception reality, and when may it not be? What is real? Is the idea itself the reality, or does it become real only when it is physically manifest?
According to M. J. van den Hoven at Erasmus University in Rotterdam, some of these issues are old wine in new bottles, but others appear to be unique to computing technology. In a real environment, ideas and physical reality are few. We can concretize our abstractions. In the near future we will have new abstractions to concretize, or we must figure out how to live abstractly in an abstract world. Lawrence Weed has said that we think in the abstract but we live in the concrete. Tomorrow this may not be true.
Unfortunately for us, as Edmund D. Pellegrino and David C. Thomasman said, ''Medicine does not have the tools within its methodology to critique its goals and values.'' We must build them as telemedicine and telematics force us to create policy to meet new realities. Our information technology policies need to reinforce the most vital roots of our moral philosophy. It is now self-evident that the marriage of information and communication technologies to health will force the IOM, government, business, journalists, and citizens to consider issues and policies that previously have been ignored, as well as to engage newly emerging realities. An example is the currently proposed legislation for confidentiality, accuracy, and the integrity of personal health records, the Bennett–Leahy bill, S. 1360. Even more is at stake, however. Similarly, telemedicine is forcing us to look at our legal structures and political boundaries.
We need to develop new methods to critique our technology, including information and communications technology. I prefer that we call this effort biotechnical ethics, in contradistinction to biomedical ethics. The focus of biotechnical ethics is upon the technology of the information communications revolution as well as other technology used by health professionals.
For example, let's seek ways to shape the design of future software, as well as our terminology, in order to achieve more moral ends than are likely to be achieved if we simply leave events to occur randomly.
With respect to terminology, consider some of the current language that is used in computing applications. Security systems are constructed with what we call "fire walls" to enhance security. Might not "privacy screen" send a clearly different cultural message and expectation? Would "house keys" rather than "password" send a different message?
Beyond these fascinating discussions and interesting word games, we must make certain that we do not pass up the opportunity to shape the future of this powerful technology, since failure to do so could be at the cost of our collective soul. While the tasks are major, the times complex, and the pace swift, the opportunities are tremendous. As John Shaw Billings stated in 1913, "there is really nothing difficult if you only begin—some people con-
template a task until it looms so big, it seems impossible, but I just begin and it gets done somehow. There would be no coral islands if the first bug [had] sat down and began to wonder how the job was to be done."
At its base, our work is honorable. Perhaps over the next 25 years, we will reinforce the best of our values and transmit them to the larger society. I hope so.
DR. SHINE: How can we get clinicians to focus upon public accountability and performance measurements?
DR. DETMER: That is a very good point because these methods are not that new. I have two stories in response. One involves the thoracic surgeons in Pennsylvania who were so disgusted two years ago by the absence of severity adjustment in reports issued by the state, that they got into the act as a society. In fact, they built a much more rigorous severity adjustment technology, presented it at the American surgical meeting last year, and essentially have owned this issue. In fact, now it is their issue, as well as being a public issue.
Similarly, it appears that in Cleveland, enough of the entire community—both providers and buyers—is paying attention to this that cost and quality are moving in the direction you would hope to see. By the way, I think Reed Gardner and the people at the LDS hospital are convinced increasingly that the very best quality care happens also to be the least expensive care.
Clearly, physicians, in order to change their behavior, must feel that data are timely and reliable, for starters. However, incentives should support this behavior as well.
I think that we are turning the corner on this, but the grand challenge is how to move from where we have been into this new era. This is doable and we are starting to see it in places, but how quickly it will spread, I don't know.
PARTICIPANT: How will confidentiality be maintained in the future?
DR. DETMER: The previous speaker mentions our living in a global village. I grew up in a village in the middle of Kansas, where everybody knew everything about everybody else. There was no such thing as privacy; you could hope for confidentiality.
I would argue that today, if people think they have privacy in the United States, they simply have not been of enough interest to the press. In many respects, actual privacy has probably always been somewhat elusive. On the other hand, I think we can greatly improve confidentiality. The basic point is
that people need to be on notice that when they are dealing with important data about someone's personal life, they have to treat it that way.
The main thing we lack at present is a national law that puts people on notice in a uniform way across the country. I think we can go a long way toward this, but some of it is where we as a society choose to come down on essentially respecting one another more.
PARTICIPANT: How will advances in information technology affect continuing medical education in the future?
DR. DETMER: Being from the University of Virginia, I will put this in a Jeffersonian context. Thomas Jefferson laid out for our country, and in many respects for the world, a system of education that was built of primary, secondary, and higher education. The assumption in that era was that once you got your basic education and went out into the world, you were prepared. You had gotten the skills and the means to go forward.
The fact of the matter is, as you saw by the growth in the knowledge base, we are increasingly reinventing ourselves on-line as we work. This country has not yet developed a strategy not just for continuing medical education but for continuing education in general. Increasingly, it becomes something our society ought to do.
I would favor a form of worker benefit that gives people so many credits for the amount of work that they do, along with the necessary educational time to get away and study.
Your question, I think, is very profound and extraordinarily relevant for a free society. As Madison also said, education literally is crucial to a free society. So it is may be a way of catching it on the wing, the way it is done now, but also developing systems that can actually help us do this. Now, clearly, information technology is going to help. A lot of people may have different views on that, but it is a key question, in my opinion.
PARTICIPANT: How will physicians use computers as "real-time" sources of information, not just in record keeping?
DR. DETMER: Gorman et al.4 published an interesting study last year showing that when doctors see patients, it is very common for questions to arise at that time, but extraordinarily uncommon for the doctors to look up the needed information. Now, if doctors have someone to do this for them or they have a decision support system, the information could be obtained while they are seeing the patient. Relevant data are found more than half of the time and, 80 percent of the time, will change what the doctor then does for the patient. We have a tremendous opportunity to leverage care if we can figure out how to access point-specific data and point-specific issues at the necessary time. However, the system has got to be lean, clean, and pretty mean.
I do not know the nursing database, but doctors typically like to get information, first and foremost, from a direct encounter with another doctor or another person. Telephones are next in line, and computers follow in terms of preference. So the broadband capability coupled with decision support systems may be helpful. It would appear that we have tremendous leverage in this.
By the way, textbooks, which are often a source to which people go, are usually sufficiently dated by the time they are published that they do not reflect what is in Medline and the databases of the National Library of Medicine, and, hence, most current expert practice.
PARTICIPANT: Where will the support come from for training the health care work force on all of the advances in information systems?
DR. DETMER: I would hope that some of the private foundations would see this as an absolutely essential investment opportunity and need, because it is part of the grand challenge. We must invest in greater education; the National Library of Medicine is doing what it can, but I think we really need to get more help on this from the foundation side as well.
PARTICIPANT: How will physicians develop trust in these new systems?
DR. DETMER: It turns out that if you have something that works and you do it right the first time, ultimately, that is best. Although that is a rather bold statement, I think there are growing examples of where it appears to play out in clinical medicine.
PARTICIPANT: Right now there seems to be a lot of marketing of services rather than sharing accurate information about outcomes. How will consumers sift through and make sense of all this information?
DR. DETMER: I think we are going to see a number of changes occur as both the media and the message move toward reality. With respect to the basic language—the terms, vocabulary, syntax, and some of those things—we are not where we could be today, whether we communicate through the Internet or any other way; and that is really what I was calling for.
PARTICIPANT: With all the focus on technology, how will issues like equity of access to technology be addressed?
DR. DETMER: This has been a worry—it is partly why I am interested in biotechnical ethics. The fact of the matter is that the technology can be sufficiently captivating and dazzling, but it can become confused as an end rather than the means to desirable ends.
If the computer interfaces are shaped properly, we will actually free up some professional time so we can deliver new treatments and learning with a human face. There is also a risk because humans are very curious and easily fascinated; we can get caught up by the glitter and think it's the goal. This
whole notion of what is a means and what is an end needs to be kept in clear focus.
PARTICIPANT: What impact will all of this additional information have on health care costs?
DR. DETMER: This is anybody's guess right now, because it is a very democratic Net out there. Anybody can throw out anything into the Internet. As I said, it can be noise or it can be signals. The problem is, of course, what all of that does. If it generates more demand for services that are off the wall and trivial, it does not help matters.
That is why I think we are calling for a new challenge to the health professional to try at least to figure out a way to address this new wave we are moving into, because with some of these devices I can see as much opportunity for mayhem as for improvement.
PARTICIPANT: It is unclear how education and research will be supported in a market-driven system of health care. How will all these advances in information technology be supported in medicine?
DR. DETMER: Looking at the current scope and scale of information technology in health care, we are underinvested within our industry by a factor of three compared to airlines, banking, finance, and such. Candidly, it is hard for me to say exactly what the effect of that will be, but it obviously could be profound. Health care is an information-driven line of work.
So as we scale up in information technology, there is absolutely no question in my mind that it will have a profound impact on how we organize ourselves. However, I really have not taken sufficient time to look at more of its organizational implications.