Enhancing Healthcare: The Significance of Longitudinal Electronic Medical Records (LEMRs)

By Jonathan D. Gold, MD MHA MSc FAMIA FHIMSS

May 6, 2024

Challenge yourself:

1. How does the transition from traditional written medical histories to LEMRs impact the efficiency and effectiveness of clinical decision-making?

2. In what ways does the graphical representation of patient data in LEMRs contribute to the understanding of temporal relationships between health events?

3. Elaborate on how LEMRs facilitate the integration of diverse data elements to provide a comprehensive view of a patient's health status.

4. What are the potential challenges associated with the adoption and implementation of LEMRs in healthcare systems, and how can these challenges be addressed to maximize the benefits of longitudinal electronic records?

The written medical history of a patient is a longitudinal record of what has happened to the patient since birth. It chronicles problems, diagnostics, and therapeutics, as well as growth milestones. By documenting events that occurred prior to the current visit, it often gives clues to current disease states and helps guide the clinician in either the diagnosis of new problems or the treatment and management of older ones.

A visual presentation of the longitudinal electronic medical record (LEMR) aims to achieve that goal by graphically displaying events plotted on a health timeline. It is a record of patient health information generated by one or more encounters in any care delivery setting, by one or more care providers. Included in this information are patient demographics, chief complaints, physical exams, review of systems, progress notes, problems, medications, plans, vital signs, history information (including past medical, surgical, medication, test, social, travel, immunization, obstetric, growth chart and developmental history), laboratory data, SOAP notes, radiology reports, genetic information, scanned documents, referral documents, and perhaps others.

The LEMR should automate and streamline the clinician's workflow through a problem-oriented or body systems approach. It can generate a complete record of a clinical patient encounter, as well as supporting other care-related activities directly or indirectly including evidence-based decision support, quality management, and outcomes reporting. Additionally, a LEMR might summarize the entire medical record but include filters and provide the user the option to determine what level of granularity is desired when accessing this view.

The important concept behind the LEMR is to show data element continuity and element relationships: show patient data points over time and show these in relation to other data points. Furthermore, an element is polymorphic: while we may have first identified element X as a chief complaint, element X might later be elevated to a patient problem and then later be considered as past medical history. An element initially declared as a medication, might be retired, and thus become medication history or may be identified as an allergy. The EMR element is important, but the lifecycle of the element given the context of events is equally important.

LEMRs can provide a general overview of the patient’s recorded medical history or can be filtered to focus on specific body systems (endocrine, cardiovascular) or specific domains (procedures, medications, etc.). They can be used to link all related information around a diagnosis or medical specialty (notes, labs, imaging, procedures, medications).

Consider three presentations from a patient’s record. The first provides the traditional text entry in a follow up note. The second is a list generated from the text using an NLP reading of the text and inference of dates (including upper and lower limits) based upon the text and metadata from the header. Potentially, if the "descriptions" (elements) were to be codified and the dates and ages captured or extrapolated, a computer-readable list with millions of data points from the complete medical record could be used to index this patient for use in predictive modeling. The third view is a rendering of that note in which temporality found in the entry has been incorporated into the patient’s health timeline. While the text record and list provide a key to the longitudinal representation, the graphic depiction of the note allows the reader to grasp key elements rapidly.

Name: Jane Doe    Date of Birth: 3/2/1948    Age: 74y

Visit Date: 7/6/2022

74 y/o white female presents with a 2-month complaint of morning headaches and nausea. She denies loss of consciousness, head trauma, or fever. Patient has a history of CHF, DM2 with chronic kidney disease-stage 2, and high blood pressure. She believes her heart failure is NY class 2. She also states she had a left hip fracture 2 years ago followed by hip replacement surgery. In 2021, she suffered a sprain of her left foot. Medication: currently taking labetalol, metformin, lisinopril, and lasix. She denies ever using insulin and is allergic to ibuprofen as it gives her hives.

Domain	Description	Time Value	Dates  point in time (lower limit—upper limit)
Header
Demographics	Date of Birth	3/2/1948	3/2/1948
	Age	74y	(3/2/1948)
Administrative	Visit Date	7/6/2022	7/6/2022
History of Present Illness
Demographics	Age	74 y/o	(3/2/1948)
Chief Complaint (CC)	Morning Headaches	2-month complaint of	5/6/2022 (4/21/2022-5/21/2022)
	Nausea	2-month complaint of	5/6/2022 (4/21/2022-5/21/2022)
Problem List (draft) (Dx)	Left Hip Fracture	2 years ago	7/6/2020 (1/6/2020-1/5/2021)
	Sprain, Left Foot	2021	7/2/2021 (1/1/2021-12/31/2021)
	High Blood Pressure	history of	<7/6/2022
	DM2 w/CKD, stage 2	history of	<7/6/2022
	Congest Ht Fail, NY 2	history of	<7/6/2022
	Allergic to Ibuprofen	history of	<7/6/2022
Surgical Hx	Hip Replacement Surgery	followed by	>7/6/2020 >(1/6/2020-1/5/2021)

Whereas the original physician note was composed through free text by the caregiver, both the list and the LEMR view have been automatically generated through NLP. When NLP summarizes the entire medical record, millions of data points get generated—some of these will prove to be redundant and many event dates will conflict. Having a clear-cut methodology to identify individual events and determine the most likely dates for these (like that proposed in the previous three posts) aids in providing a near-true record of the patient’s status. While a constantly changing, dynamic, and inclusive list rapidly becomes unmanageable for the human reader to comprehend, it allows the patient’s data to be used by AI to index him/her for predictive modeling, decision support and other purposes.

Conclusion

The written medical history of a patient serves as a crucial repository of the patient's health journey, offering invaluable insights for clinicians in diagnosing and managing current conditions. Incorporation of a Longitudinal Electronic Medical Record (LEMR) view accentuates the narrative captured in free text, organizing and highlighting key components. LEMRs provide a comprehensive overview of a patient's medical history, facilitating efficient decision-making and improving patient care. The integration of various data elements into LEMRs, including demographics, clinical notes, laboratory results, and imaging reports, enables a holistic understanding of the patient's health status over time. Moreover, the dynamic nature of LEMRs ensures that patient data is continuously updated and contextualized, reflecting the evolving nature of the patient’s health status. As healthcare providers strive to enhance patient outcomes and streamline workflows, LEMRs emerge as a transformative tool, empowering clinicians with actionable insights and enabling personalized care delivery.