Donald A.B. Lindberg: A Digital Pioneer at the National Library of Medicine
By Peter Wortsman
A vintage documentary film clip from 1966 shows an introverted young researcher at the University of Missouri taking pains to contain his excitement. Having devised the worlds first automated laboratory system before the field he helped create had a name, Donald A.B. Lindberg58 proceeded to announce that it was now possible for the computer itself to discover new patterns and new syndromes in the information about patients. The age of medical informatics had begun and medicine would never be the same.
A Romance Between Physicians and Computers
Two years earlier, in a visionary paper titled, A Computer in Medicine, Dr. Lindberg predicted the development of a romance between physicians and computers which will rival that seen with stethoscopes and electrocardiography. It will be a great but pleasant challenge, he added, to see that this latest affair of the heart makes a significant contribution to medical knowledge and wisdom.
The romance Dr. Lindberg predicted and fomented has indeed blossomed into a virtual marriage between PCs and MDs. Modems have put the latest findings and data at a physicians fingertips. The new frontier of molecular research would be unthinkable without computers to store and process genetic data. Automated laboratories, hospital diagnostic equipment like PET and MRI, and high-tech classrooms, lecture halls, and libraries have fundamentally changed the teaching and practice of medicine in our time.
His is an astounding career that took him from Columbia, Mo., where he straddled the disciplines of pathology and information science for more than two decades, to Washington, D.C., where he has served since 1984 as director of the National Library of Medicine. From 1992 to 1995 he helped create and run the National Coordination Office for High Performance Computing and Communications.
A tall man, trim and energetic, his youthful enthusiasm unabated at age 65, Dr. Lindberg exudes the earnest charm of a Jimmy Stewart in Mr. Smith Goes to Washington. The difference, of course, is that he isnt acting. I was warned about the federal government, Dr. Lindberg likes to joke, but I correctly suspected that 20 years in a state university was adequate training for any amount of bureaucracy they could muster around here.
Ably and eloquently communicating his dynamic vision of the National Library of Medicine to Congress, the keepers of the purse strings of progress, he has managed in a decade and a half at the helm to oversee the Librarys transformation from a distinguished, albeit dusty, repository of medical books and papers to a cutting-edge electronic clearinghouse of biomedical knowledge.
Housed in a fittingly modernist complex on the campus of the National Institutes of Health in Bethesda, Md., the Library is Dr. Lindbergs laboratory where vast holdings include GenBank, the NIH database of DNA sequences, the very existence of which explodes our notion of a library. His object, as he put it in the 1993 Dainton Lecture at the British Library, aptly titled High Performance Libraries, is to remove barriers that stand between library users and the information and knowledge they need. To that end, he has been busy building and improving technical, organizational, and conceptual bridges among the Library, health professionals, and the public.
Programs devised and refined under his direction include, among others, Grateful Med and its updated Internet version, access interfaces to MEDLINE (electronic heir to the famous Index Medicus, the worlds largest collection of published medical information), and Unified Medical Language Systems, an electronic thesaurus to medical terminology. Diverse databases in such specialty areas as toxicology, cancer, bioethics, and health care administration also fall under Dr. Lindbergs leadership. Another startling innovation is the Visible Human, a human database comprising a 3-D set of sub-millimeter resolution points of an adult male and female with countless potential applications, including virtual reality teaching and remote telesurgery.
The medical profession just needs a little bit of a chance to get used to some of these new tools and be ready to make use of them, says Dr. Lindberg. The question, he adds, is not, as some asked in the past, Can the computer do better than the human? but rather, Can the computer and the person do better than the person alone?
A Pathologist with a Binary Bent
Earning his B.A. degree in biology (magna cum laude) from Amherst in 1954, back in the high-tech dark ages when business machines were still big bulky monsters of little practical use, Donald Lindberg gobbled up every subject in his path at P&S. I fell in love with every medical specialty, he recalls. I was going to become the worlds greatest surgeon, internist, obstetrician, psychiatrist, you name it, but I ultimately decided that there were probably more answers in pathology, and I was after answers!
At P&S, he benefited from the encouragement and support of Harry Rose and Yale Kneeland, who graciously invited him to join their daily infectious disease rounds, and pathology chairman Harry Pratt Smith, who provided him with the laboratory space and funding to launch his study of gram-negative pneumonia, later earning NIH support. The challenge of a good medical school, says Dr. Lindberg, is to provide an atmosphere in which the student, who wants to learn everything, can come pretty close to that but not become so stressed as to squelch his or her own creative juices.
Another profound intellectual influence at P&S was Robert Loeb: He could take a seemingly very ordinary patient with a very common ailment and make that individual the worlds most fascinating creature, whose every electrolyte value was alarming and amazing. While diagnostic artificial intelligence programs based on an encyclopedic approach have been developed, Dr. Lindberg has set as his ultimate, and as yet unfulfilled challenge, creating an electronic model of medical diagnosis that reasons from basic principles, the way Loeb did.
Columbia also did him the ultimate favor. It was at P&S that he met his wife, Mary Musick, then a nurse in the pediatric service of the Vanderbilt Clinic.
Pursuing an internship and residency in pathology at Presbyterian Hospital, Dr. Lindberg took an NIH research grant along with him to the University of Missouri in 1960, where he started as a resident and later joined the pathology faculty and directed the medical centers Diagnostic Microbiology Laboratory.
The bucolic setting of Columbia, Mo., where his three sons were born and where he and his family liked to ride horses in their spare time, provided fertile ground for intellectual inquiry. While continuing his research (rising in the academic ranks to professor of pathology in 1969), he found it increasingly difficult to balance his time among running the microbiology lab, teaching, and conducting his NIH-funded research. Then one day, I finally figured out that information access was at the root of it all.
A Computer Epiphany
It was, as he likes to put it, an idea born of bourbon and branch water. Actually, the idea emerged in the course of a social engagement with a physics student and his wife, a lab technician. Before the evening was over, we saw no reason why we couldnt build a machine to look at the antibiotic sensitivity of bacteria growing in liquid media. It took us a couple of years, but we did it.
His early prototype displayed results by lighting up numbers on a vacuum tube. At first he wrote down the results by hand. He later graduated to using a Victor adding machine, pasting up the strips of tape. Finally, we discovered that there was a thing called a computer that would save all the time and trouble of cutting and pasting. The computer was a Burroughs 205 located in the math department and he had access for 15 minutes each night, from midnight to 12:15 a.m. The computer permitted him to actually make mathematical models of bacterial growth.
And once I realized that the problem I was having with this lab was soluble by an information processing system, he recalls, it suddenly dawned on me that, heck, a lot of other problems would be soluble in the same way. Then I finally fathomed that you could follow a single patient, use the computer to store and recap all that information on his work-up. I was off and running! I wanted the patient record to be the subject of research and to get the computer to do it. The result was the first automated lab system in the world. It was one of those things that couldnt be done--and certainly not in Missouri--and, by George, we showed em we could!
Subsequently asked to join an NIH study section on computers and biomathematics, he was quickly disabused of any incipient hubris. It was the equivalent of a four-year postdoc fellowship with people who knew much more than I did about computing, mathematics, and engineering.
Meanwhile, he began publishing articles in a field that would come to be called medical informatics and by the mid-60s he had developed an international reputation. In 1967 he became director of the Missouri Regional Medical Program Information Systems. In 1969 he was named professor and chairman of the Department of Information Science in the School of Library and Information Science.
Then in 1984 came the call to the capital. Despite the difficulty of pulling up roots from a town he and his family had come to dearly love, he quickly took to Washington ways.
Dr. Lindberg credits the sound political assistance and counsel of such farsighted politicians as Sen. (and later Congressman) Claude Pepper in helping to communicate the Librarys message to Congress. It became quite clear in preparing for congressional hearings, that if we went and talked about how far behind we were in cataloguing books and articles or shelving dusty memoirs, we wouldnt get anywhere. But we could--and did--produce a string of patients who had been treated successfully with drugs produced by the recombinant DNA process, thanks to the NLM. He managed to convince Congress that the Library was a conduit to information essential in the decision-making process of scientists and pharmaceutical firms and that the ultimate beneficiary of such information access was the patient and the general public.
High Performance Computing
In 1991, then Sen. Al Gore wrote and shepherded a bill which, with bipartisan support, was to become the High Performance Computing Act. Dr. Lindbergs informatic credentials and experience came to the attention of the executive branch. In 1992, when a call came inviting him to lunch at the White House, he remembers remarking to an associate: Were really in trouble now. Something has gone sour. Theyre hauling me in for a dressing down at the Oval Office! President Bush appointed him director of the newly created National Coordination Office for High Performance Computing and Communications.
Completing his HPCC tenure in 1995, he looks back on a job well done. It was a lot of fun and I learned a lot. It had nothing per se to do with medicine, and thats why I agreed to take on the challenge. I realized that unless somebody with some understanding of biomedical concerns was heavily involved right at the outset, medicine would be simply bypassed.
In 1996 Dr. Lindberg was named U.S. National Coordinator for the G-7 Global Healthcare Applications Projects, a post he still holds along with his Library responsibilities.
A member of the Institute of Medicine of the National Academy of Sciences, he has been showered with honors and awards, including the Surgeon Generals Medallion of the U.S. Public Health Service, the Presidential Rank Award of Meritorious Executive in the Senior Executive Service, and the prestigious Morris F. Collen, M.D., Award of Excellence of the American College of Medical Informatics.
Dr. Lindberg holds honorary degrees from Amherst, the State University of New York Health Science Center at Syracuse, and the University of Missouri. He has lectured worldwide.
A proponent of Thomas Wests view of the computer as a tool to help foster a new generation of visual thinkers-- Apparently Einstein would first see the answer, then he would find words and numbers to express it.--Lindberg posits: Is it not a delightful thought to imagine such a master with a modern computer. And why not, one might well add, imagine a generation of biomedical wizards first picturing and then unlocking the mysteries of the human genome, thanks in part to the pioneering work of Donald Lindberg.