In the NNLM Big Data in Healthcare: Exploring Emerging Roles course, we asked participants, as they progressed through the course to consider the following questions: Do you think health sciences librarians should get involved with big data in healthcare? Where should librarians get involved, if you think they should? If you think they should not, explain why. You may also combine a “should/should not” approach if you would like to argue both sides. NNLM will feature responses from different participants over the coming weeks.
Written By Beth Whipple, Assistant Director for Research and Translational Sciences at the Ruth Lilly Medical Library, Indiana University School of Medicine
Big data is one of the directions in which the field of healthcare is moving, and to continue to support and collaborate with our colleagues outside of the library, we need to understand trends and how to provide relevant resources and support. As experts in information retrieval, information organization, and as folks who interface both with end users and back end developers, we are uniquely positioned to be involved with big data in healthcare. I see roles for health sciences librarians in four general areas: programming/coding, information organization, end-user/usability feedback on systems, and data management.
As an undergraduate math major (who also had to take computer programming classes), I find it interesting to see how my previous training now relates to what librarians are starting to do, in particular the involvement of some data librarians in programming/coding instruction (e.g., teaching R, Python). That being said, there is a reason I went to library school and did not get an advanced degree in math. While librarians can build roles in this area, I believe it is not for everyone, and there are other ways that librarians can be involved in big data and data science work in healthcare.
Information organization is a big area where I see librarians involved with big data moving forward. While we are most familiar with literature databases, I often explain to patrons that if they understand how one database is set up, they can use those organizational principles to understand other databases. For example, as part of an NLM Informationist project at my institution, three librarians created a map of all the rules for a clinical decision support system to show how items were connected and to identify gaps. While we did have to learn how to read through the rule syntax, which presented a learning curve, we really were using our information organization skills to create maps of different concept areas and visually present that information to the pediatricians we partnered with on the project. The clinicians looked to us for expertise in the area of information organization in order to better understand their clinical decision support system.
The third area in which we can contribute related to big data is through our end-user and usability skills with our patrons and clients in how systems are designed. We are familiar with straddling the line between understanding the technical side of systems and translating them to our users. I also sometimes see our expertise acting as a squeaky wheel to try and explain to technical folks why something they think is “so cool” isn’t 1) practical, 2) useful, or 3) necessary. As a knitter, just because there are many things that I could make, doesn’t mean I should. Sometimes designers can get carried away with something technically interesting that is totally useless. Our role in that instance is to speak up, reiterate the desired outcomes of the project, and help make sure the end goal is reached.
The fourth area we can provide support for big data is through data management. I taught a Tableau class yesterday, and in the debrief with my colleagues, it was pointed out that I was teaching data management without even realizing it. As part of the class, I pointed out a sample dataset’s naming conventions and mentioned that those outside the project might not understand those conventions. I highlighted the importance of considering naming conventions when working with datasets, in order to ensure clarity. Additionally, my Data Services Librarian colleague related recently how, in working with our Clinical Informationist, she learned that he keeps a “diary” for each systematic review he’s involved with where he records details about the search strategy, databases searched, and documents other pieces of the review process. She talked with him about that practice being a form of data management, which hadn’t occurred to him previously. Many librarians are already practicing data management and teaching those skills in their everyday work, without realizing it’s “data management”. Librarians can easily expand their roles to support big data through this area, as information organization skills are underlying aspects of big data and librarianship.
As health sciences librarians, we are connectors – helping to bring the right people together, leading the right people to the right resources, and bridging the gaps between silos. We can demonstrate this through offering classes at the library – taught by library staff or other experts – on data topics, sponsoring data talks through the library, and in general doing what we do best—serving all patrons that are part of the mission of our institutions, sharing information, and connecting people, in order to make things more efficient and productive overall.