Sunday, June 15, 2014

National Library of Medicine: Past, Present, and Future

This week I am off to a meeting I look forward to every year, the annual meeting for trainees funded by the National Library of Medicine (NLM) biomedical informatics research training grant program. This meeting has been valuable for its content and networking back to the time I first attended it as a trainee in 1988. I wrote about it in the past in this blog.

The NLM is truly an exemplar not only for the United States (which funds it via Congress), but also the rest of the world as well as many people individually, including myself. The NLM is unparalleled in providing access to biomedical and health knowledge. It not only provides operational systems used around the world daily but also has a robust research and development program that develops, implements, and evaluates the next generation of biomedical and health knowledge tools. The NLM provides access not only to clinical information, but also biological, consumer-oriented, and public health data and information. As one of the institutes of the National Institutes of Health (NIH), it adds value to the research missions of all the other institutes in an integrative manner.

Starting with my postdoctoral fellowship from 1987-1990, the NLM has certainly enabled success in my career. It also funded my first research grant, one of the old First Independent Research Support and Transition (FIRST) or R29 awards, and then numerous other research as well as training grants. The NLM also funded the initial program and the building where my office resides under the Integrated Advanced Information Management Systems (IAIMS) initiative at Oregon Health & Science University. Like many in the informatics field, I owe a great deal to the NLM for my accomplishments.

At this time, the NLM is preparing for its next round of its long-term planning. This process is slated to begin in 2015 and will result in a publication of its next long-term plan in 2016. This new plan will supersede the last long-term plan that was published in 2006.

In preparation for the next round of long-term planning, the NLM recently held a symposium to reflect over the 30 years of leadership of its current director, Dr. Donald AB Lindberg. The Web site for the symposium allows visitors to leave comments about the past, present, and future of the NLM. The latter comments will be used among the information-gathering efforts to launch the next long-range planning process next year. What an opportunity for the Informatics Professor to share his thoughts on the future of the NLM, which I will put in the rest of this posting and then paste into the future comments portion of the NLM symposium site.

In light of the NLM's success, what recommendations do I suggest for moving into the future?

One of the big challenges for the NLM is its name. I am not one of those people who believes that "libraries" are something of the past. I have a whole chapter in my book on information retrieval (search) devoted to "digital libraries" and their importance [1]. Libraries are certainly changed in the 21st century, but still remain repositories of data, information, and knowledge, even if much of it is now digital. It is still important to have libraries and librarians who collect, curate, make accessible, archive, and preserve data, information, and knowledge. The fact that its materials are mostly digital now does not eliminate the need for these other functions.

However, the NLM always has been more than a library and will likely continue to be so. The NLM funds intramural (within in NLM) and extramural (outside NLM, mainly in universities) research, mostly in informatics. It also provides and funds education, mainly to librarians and future informatics researchers, but also many others. What the NLM really does then, in the big picture, is biomedical and health informatics.

Since the NLM is one of the institutes of the NIH, we may ask then, why is the NLM not called something like (my preferred name) the National Institute for Biomedical and Health Informatics? This would not only reflect its larger activities beyond being a library, but also make its role more clear, especially for those to whom NIH research funding is important (such as Promotion and Tenure Committees at universities). If this name change were made, the library function of NLM could be one of its major divisions, synergizing with its other informatics functions.

Another important recommendation for the future is to expand informatics research funding. If for no other reason, this should be done to give informatics research a fairer share of NIH research funding. Is informatics research really only deserving one-tenth or one-hundreth the funding of the major disease-based institutes? I know that substantial expansion of overall funding is unlikely to come to the NIH any time soon, but consideration should be given to re-prioritizing the role of information systems and technology in achieving the "triple aim" of healthcare (better health, better care, and lower cost) [2].

In teaching to various audiences, I have noted an analysis by Woolf and Benson comparing the overall health benefit of more people having access to a treatment versus the incremental improvement in treatment efficacy [3]. This analysis is mainly used to make the case for investments in universal healthcare (something with which I agree), but can also be applied to the notion that informatics research helps us understand how to provide better access to patient data, along with information and knowledge to deliver more effective care.

I also would advocate that the NLM (or this new institute) should expand its research explicitly in clinical informatics. I argued recently in this blog that the NIH informatics research agenda needed to expand its focus on the role that data, information, and knowledge play in complex clinical settings. I also expressed concern that data science not taking into account the context of health and healthcare might not achieve its potential.

In fact, this newly renamed institute could also be the home of other current activities, such as data science, which is so dependent on informatics and would also synergize with its library function. At a time when many organizations are developing matrixed organizations where the different entities perform integrative functions, NLM can and should take on that role within NIH, with its core function centered around informatics.


1. Hersh, WR (2009). Information Retrieval: A Health and Biomedical Perspective (3rd Edition). New York, NY, Springer.
2. Berwick, DM, Nolan, TW, et al. (2008). The triple aim: care, health, and cost. Health Affairs. 27: 759-769.
3. Woolf, SH and Johnson, RE (2005). The break-even point: when medical advances are less important than improving the fidelity with which they are delivered. Annals of Family Medicine. 3: 545-552.


  1. Your posts are personally highly regarded and always insightful. I have believed that information technology would improve everyone's lives for sometime now as the Special Interest Groups on Information Retrieval & Technology Education of the Association of Computing Machinery) have advanced these areas of interest and relevance to Health Care Informatics for decades now. I hope you enjoy your NLM Bioinformatics meeting and post sharable information with your audience (of which I have become an avid follower). I presume you no longer are involved in direct clinical patient care as your other responsibilities appear highly time-consuming. Best wishes!

  2. Bill, I agree with you on the future of the NLM and hope that you have a say in the next steps!

  3. Thank you, Chris. I visited your ONC-HIT video post and found it helpful, although it is "singing to the choir", in my case. I agree that HIT can improve delivery of care and ease the burden on providers (once trained). Dr. Hersh's blog on the NLM is also fully supported by me and and experience as it has been a valuable source of health related information. My university library provided me with free full-length articles when the NLM PubMed site provided only abstracts. However, with the increased need for Evidence Based Medical Practice research studies additional sites should be consulted. The controversy over "cookbook" medicine needs to be shelved and a serious implementation of the EBP (evidence based practice) discipline promulgated so that better and clinical applicable findings can be used to better patient outcomes. Again, thanks for your support and let us serve as champions of health care informatics along with Dr. Bill Hersh!