Skip all navigation and go to page content

SEA Currents

Newsletter of the NN/LM Southeastern/Atlantic Region

An electronic medical record - system (EMR-S): a first-hand experience

Prepared By: PJ Grier, Outreach/Access Coordinator, NN/LM, SE/A

Various viewpoints abound about the wisdom of using physician shadowing as a learning tool for students.  One end of the range supports a student’s need to experience medicine’s practical side[1] while the opposite end holds physicians accountable for ethical obligations to their patients[2]. A middle ground exists, somewhere, because medical schools continue to use the tool as a vehicle to foster student pursuit of the profession.  With prior written agreement and patient acceptance a shadowing event must be grounded in patient privacy and confidentiality, and access to patients must be provider supervised at all times[3].  As a medical librarian my purpose is to use shadowing as a gateway for gathering physician-centered data from University of Maryland Medical System (UMMS) ambulatory clinics at various stages of Portfolio Ambulatory©[4], an electronic medical record system, implementation.  Though physician shadowing has traditionally been used to inspire students to pursue the field of medicine, shadowing may have larger research implications for health science information professionals, faculty and students pursuing a health sciences specialization from a library science program.

As someone with a keen interest in clinical informatics, more specifically the interplay of various terminology standards, sharing a few observations on electronic medical record – systems (EMR-S) and medical librarianship seems worthwhile.  Gladly, this is a welcome by-product of recent independent research for certification completion in clinical informatics from Johns Hopkins.

Medical librarians can improve upon an EMR-S’s human computer interaction (HCI) factors, from development, design and implementation standpoints. Because there are many flavors and capabilities of EMR-Ss (ambulatory, hospital, specialty-based, etc.), HCI efforts are usually encoded in the R&D portion of an EMR-S lifecycle and tweaked or customized during end-user implementations and deployments. An example of an HCI issue is “patient mismatches” meaning provider narrative data has been entered in an incorrect patient record. Narrative data is where the physician enters (either through a keyboard, or digital dictation) the patient’s objective/subjective assessment for chief complaint, history, diagnosis, treatment and assessment of tests. Since no two patients are ever alike, it follows that clinical narratives also are not exactly alike. The content of the patient narrative is really the “core” of any EMR-S and should be of interest to medical librarians. If the EMR-S does not have an effective process to counter patient mismatches, it can become a barrier to improved patient quality and safety.

Health science librarians should know about the “info-button”, an HL7 compliant feature which when activated allows an EMR-S user to link-to, or “jump-to” a clinical e-resource for published literature discovery. Future EMR-S developments are to make functions transparent, particularly when they involve billing and reimbursements.  Contributing to this push is that ICD-10 clinical documentation will be required by 2013. Here again, the clinical narrative is front and center. While EpicCareÓ[5] employs documentation tools to assist the physician in completing the patient narrative, the narrative’s content still contains mostly unstructured data. While not yet perfected, computer synthesis of unstructured narrative data is essential for billing, revenues/reimbursements as well as analytics including surveillance, research, outcomes, and performance management. In order for successful computer synthesis to occur, it requires the maturation of natural language processing (NLP), artificial intelligence (AI) and pattern recognition techniques coupled with the integration of medical terminologies (i.e.: SNOMED-CT, LOINC, etc.) and domain thesauri; all of which are supported by behind-the-scenes computer processes that cross-walk, auto-suggest, correlate, and harvest the semantic representations of the physician’s written words in the narrative.  This is just to say, that, health science librarians having skills or even a penchant for managing vocabularies, thesauri, and standard terminologies can make professional contributions.

Commercial interests are looking to exploit computer synthesis of the narrative primarily driven by healthcare reform and the need for improved healthcare coding mechanisms. Remember what I said earlier that the “physician narrative should be of interest to librarians?” It is because of another important healthcare opportunity that could ignite the further adoption of NLP techniques in scholarly publishing.  Once the developmental science of NLP and it cousins are more perfected, there is every reason for a publisher to want to integrate appropriate electronic literature on a patient’s diagnosis and treatment directly within the EMR-S’s narrative, where it ultimately belongs. Accomplishing this concept is fraught with pitfalls, such as radically changing the publisher-pricing paradigm and possible impacts to health science librarianship, as well.  However it is only a matter of time before the developmental science of NLP catches up, thus opening the potential to make published literature discovery more transparent than through an EMR-S’s info-button.

If these topics are interesting, then I invite further exploration in a newly designed class called Informatics for librarians – peeling the onion. The SE/A co-developed class is offered in-person and comes with 3MLA CE upon completion.


[1] Wong KR, Gold, JA. Letter. JAMA. 2011;2114-5.

[2] Kitsis EA. Shining a light on shadowing. JAMA. 2011;1029-30.

[3] University of Washington. (Internet). Application and agreement for shadowing, and/or observation. (cited 2011 Aug 9). Available from

[4] Portfolio© Ambulatory is the service mark of the University of Maryland Medical System (UMMS) for its EpicCare Electronic Medical Record system (EMR-S)

[5] EpicCareÓ is copyright of Epic Systems Corporation, 2011

Comments are closed.

Last updated on Tuesday, Nov 8, 2016

Funded under cooperative agreement number UG4LM012340 with the University of Maryland, Health Sciences and Human Services Library, and awarded by the DHHS, NIH, National Library of Medicine.