I write this blog for a number of reasons.
Sharing my ideas helps me clarify my thoughts. It helps me plant a seed to advance an idea or enlist your support. It also gives me the opportunity to introduce different aspects of the National Library of Medicine and its vast array of services.
And obviously, I write each post intending it to be read, but who exactly are my readers?
A coach once told me that, before I start any writing project, I must envision my readers. Here are the readers I envision right now:
Some of you are stakeholders of the Library and its products and services—researchers, clinicians, librarians, patrons, policy makers, and NLM staff members. Many of you have already shared your ideas, stimulated by some of my writings; I encourage more of you to do so!
Other readers are my loyal friends and professional colleagues who want to know what I’m up to in this phase of my life. These folks are more likely to tell me privately they’ve read the blog or they look forward to the next installment.
Still others find their way here through the power of connections and the exhortation to “read this.” That is, they might get a link to the blog post from someone who found an idea to be interesting, provocative, or maybe even wrong.
Wrong? Could there be someone who takes issue with or maybe even disagrees with a perspective I have advanced?
Of course. Probably a few someones.
I recognize that people may (and likely do) hold different perspectives on some of the ideas I’ve already advanced, such as providing information to people on the move or the sly bemusement I expressed regarding the Cold-War origins of the building. Sometimes I even hear from them. That’s why there’s space here for readers to talk back.
Bring it on, I say!
Because ultimately, the reason I write this blog is to engage with you.
What’s on your mind? You have the floor.
Hi Dr. Brennan. I listened to your talk about data and NIH. It is interesting to me when you mentioned training librarians about data, and also training clinicians on how to use data to inform patient care (more than just EBP) but not necessarily training them to be data scientists. As a librarian and a clinician, I would be interested to see how that would be done. What would be the content? Given the amount of information the clinician needs to know to provide patient care, would we have the time to incorporate this into our usual routine? Thanks for your willingness to listen.
Hi, Frances.
Thank you for your message and for listening to the talk. Others who are interested can watch a 25 min video of the talk here (starts at time 50: 00): https://videocast.nih.gov/summary.asp?Live=21783&bhcp=1
For all students, it is critical that the fundamental concepts of data science be woven through the library science and health professions education, through research courses (as methodological examples) and through substantive courses, as examples in support of core concepts. In addition, I believe that at least two courses would be necessary for the data science librarians and two for the data-informed clinicians. A foundational course in data science, including philosophy, methods and examples, should introduce health data science to any graduate students interested in this specialization–it would indeed be an excellent opportunity for inter-professional education. For the LIS curriculum, a second course should build on the core LIS curriculum, applying key concepts of acquisition, curation, and management to large data sets, along with data discovery, patron services, and techniques for data deposit involving public and privately-held large data sets as workspaces. For the clinician who seeks additional training in data science, the second-level course should cover generation of knowledge through large data set analysis, engaging patients in data discovery, and critical appraisal of the evidence provided for clinical care from data-driven discovery projects.
Of course, it’s early in our exploration of data science as a foundation for discovery and clinical practice and there may be lots of other ideas — please share them!
Best,
Patti
I read your blog because I am a hospital librarian and I rely on the NLM in my daily work. I like being informed and I appreciate the opportunity to guide the direction of the NLM as one of many stakeholders.
Hi, Kathy —
I am so glad to hear from a hospital librarian! Your part of the profession plays an essential role ensuring that the resources of the NLM inform care in a timely fashion. We look forward to continuing to hear from you as we craft a vision for our third century!
Best
patti
What’s on my mind:
data science topics such as
– Common Data Models for large datasets of EHR data
– How terminologies support better datasets (that can answer more questions or with less bias)
– How to show the value of healthcare bigdata to a healthcare consumer (produce products that help answer consumer health questions) (average citizen out there)
– Relate research CDEs to routine healthcare data
-How we match researchers to datasets out there (e.g., licensed claims data, networks with EHR data, large sites with EHR data (e.g., KaiserPermanente in CA), registries out there), feasibility assessment of a retrospective healthcare bigdata research question
Vojtech Huser
Vojtech —
Great to hear from an NLM colleague here! I am in sync with your ideas: Models and terminologies produce interoperable models of health data, which in turn ensure that we can glean value out of operations. Some of the questions that you raise are research questions (like how Common Data Elements relate to the everyday health experience — I think that CDE’s are optimized for research, not sure they have the same clinical value) and some are just hard engineering research questions — like linking large data sets!
Great ideas. Glad you are reading!
patti
I read your blog because I’m curious about the direction that the National Library of Medicine is taking under new leadership. I’m an assistant director of one of the National Network of Libraries of Medicine’s Regional Medical Libraries, out here in the Midwest, and I’m new to the organization and have a lot of questions. I want to learn about how our work supports initiatives at NLM and how we can engage the health sciences libraries, public libraries, and other health information agencies in our region to take part.
Hi, Elizabeth —
It’s great to meet one of the NNLM RML staff — how’s that for acronyms!
The NNLM will continue to play important roles in assuring the health of the public as they always have, and will take on new responsibilities as we move towards a data driven society in an era of precision medicine. The librarian has a responsibility to the public to stay up to date, to meet patron needs, and to help them access relevant and actionable information. As health information becomes more data driven, librarians will serve as the interface to many publics. It will be important to help clinicians and scientist access the literature and know how to interpret the results of data science-based studies, along with the more familiar experimental studies. Some RML/NNLM librarians will actually help investigators locate and access data sets, or upload their data sets into large data repositories. These librarians can also be in place to help patrons locate and understand clinical research studies they might participate in, interpret ideas they hear in the news, and discern safe and reliable sources for health information.
Sounds like what librarians have always done — just now complementing literature with data!
Best,
Patti
I read your blog to try to learn what your interests are re the future of libraries in general and NLM in particular. My hope is that it will help me understand where and how the work I currently do fits in with the paths you want NLM to take going forward, and therefore how my work here might evolve to meet those goals.
As a librarian whose position is entirely about working with different kinds of data in different kinds of environments and is deeply involved in the evolution of modeling and exposing library data, along with use, re-use, and management of library data, the comment above about ‘training librarians about data’ is of interest but probably needs context. I don’t know the source/which presentation is being referred to, but can you expand on that comment?
Hi, Nancy — and greetings to the NLM LO colleagues.
The comment in the previous post refers to remarks I made during the NLM Board of Regents meeting on 2/7/17. You can watch a 25-min video of the talk here (starts at time 50:00) — https://videocast.nih.gov/summary.asp?Live=21783&bhcp=1
I think a lot about how the Library will foster data driven discovery in the future — and know that we must both expand the patron services we provide as well develop new methodologies for curation and collection management.
I would be interested in hearing your thoughts!
Best,
Patti
I am the Information Technologies Librarian at the Alabama College of Osteopathic Medicine. I used to work with the National Network of Libraries of Medicine, Southeastern Atlantic Region, and I read your blog because I am interested in the future of the NLM and wish to provide updates and information to colleagues and researchers here.
Hi, Shelia — It’s so great to hear that people in practice view the NLM as a support to colleagues and researchers. Please keep reading and let me know what other topics you’d like to hear about!
Best,
patti
I read the blog out of interest in where things will go with the Data science plans, as well as interest in what tools and techniques will be brought to bear to get the right piece of information into the right hands at the right time. I was very interested by your initial speech about your vision for NLM and signed up for notices when you started the blog. I’m a data plumber, designing content and metadata flow, tracking provenance and integrity, and deriving new connections and knowledge. I look forward to learning more about what standards you will choose or create for datasets, for provenance, for linking, and more.
Love the job title “data plumber”! There are all sorts of new skills needed in this new era. I want to pick up on your use of the term provenance, and linking it with integrity — right on! Provenance — the ability to know the origin and history of a data element or data sets, including how they have been transformed, analyzed and interpreted — is a key contributor to reproducibility and rigor, the hallmarks of high-level science. Indeed, we need strong computational tools to help demonstrate and manage provenance.
Best of luck with your work!
Patti
Hi Patti.
Maybe you’ll be interested to know that I am a flight attendant for a major airline, flying internationally out of the NYC area. I’m a mom with two teen boys, and I’m scared for their future due to the fact that our country is failing to act on the urgency of climate change. I read your posts because I try to support scientific research in all regards and I yearn for the day that information silos are breached so that the American public can start to connect the dots between our extraction/consumption paradigm and our health. I believe that not only are we killing our current and future generations with our toxic waste but also our very planet. I appreciate your blog for many reasons, not the least of which is that you make those of us who are not scientists feel welcome. I have utilized the Library for research on cancers that have plagued both of my parents, both of my in-laws, and many friends. The up-to-date information gives me hope in the midst of misery. Thanksfor asking.
Many Thanks for your reply! I sent a brief note to your email asking to participate in the conversation if approaches for tracking provenance and maintaining integrity are discussed by your teams.
Beverly
Hi Patricia
Meanwhile, back at the ranch……. two really good things the NLM has done: making the original Forgotten Frontier film and the Zwerdling postcard collection available online. World class and absolutely brilliant.