The Effect of Technology on the Written Tradition of Medicine

THE EFFECT OF TECHNOLOGY ON THE WRITTEN TRADITION OF MEDICINE BRYAN P. BERGERON* Medicine is a complex discipline requiring an ever-increasing vocabulary to communicate and record concepts for clinical, research, and reimbursement purposes. Because of time pressures and, increasingly, economic constraints , a plethora of abbreviations—a medical shorthand of sorts—has developed over the years. For example, when recording patient history by hand, virtually all clinicians use the shorthand "y/o" for "year old"—a habit learned in medical school. This medical shorthand has been cataloged at various stages of development [1], but because of the rapid rate of technologic innovation in each specialty, there is considerable overlap and frequent additions. As a result, many medical records are fraught with ambiguity and lack standardization, effectively isolating readers outside of narrow subspecialties or geographic regions. As the medical community moves from a handwritten paper chart to an electronic medical record (EMR), the differences between handwritten, transcribed, and computer-based communications take on practical significance . The healthcare industry is increasingly looking to the EMR to control costs, increase communications efficiencies, and provide more effective patient care [2] . Thanks to the growing information infrastructure, EMR data potentially can be shared instantly with medical colleagues in multiple institutions, quality assurance and outcomes researchers, and insurance and government agencies. Once the data in the medical record are in digital form, they can be searched, reformatted, transmitted, and archived with relative ease. To the dismay of many clinicians, the driving force behind the EMR is primarily economic. Paradoxically, given the stiff record-keeping requirements for quality assurance and reimbursement, there is even more demand for personal record -keeping systems to assist in practical clinical record keeping. Often, this consists of little more than patient ID, diagnosis, medicines, age, and *Medical Informatics Laboratory, Department of Anesthesia and Critical Care, Massachusetts General Hospital, Fruit Street, Boston, MA 02114.© 1998 by The University of Chicago. All rights reserved. 0031-5982/98/4104-1068$01.00 572 Bryan P. Bergeron ¦ Written Tradition ofMedicine a short to-do list on a 3 X 5 card. Often this working document never finds its way to the official medical record, whether handwritten or electronic. With this dual record-keeping clinical system, there's even more time pressure to document clinical status as quickly as possible. Also, when a document is limited to a 3 X 5 card, abréviations are quicker and more space efficient. Some EMR systems, mainly those developed by clinicians, place clinical record-keeping capabilities at the forefront of the medical record. However , this paradigm seems to be successful only mainly for small clinics and private practices. In larger clinics, hospitals, and healthcare organizations, administration, accounting, information services, quality assurance, and other departments typically manage to define an information system that places the clinician's needs somewhere near lowest priority. Although creating an effective and clinically useful EMR system presents a number of political, economic, technical, user acceptance, and legal issues , data capture and access typically present the greatest challenges of all [3]. For example, there is the issue of how to best support the clinician data entry process, from the use of the keyboard, mouse and pen, to voice recognition. Another issue is that of using a standard vocabulary, such as the National Library of Medicine's Medical Subject Headings to facilitate search and retrieval, automated decision support, data mining, and quality assurance measures [4, 5]. Standard vocabularies provide a consistent concept mapping that controls for differences in regional and national medical practice, and therefore allows for relatively rapid comparison and searching of medical records from a variety of sources. For example, without a controlled vocabulary, the finding of acute myocardial infarction could be recorded by one clinician as "Acute MI," by a second as "MI," and by a third as "Acute Myocardial Infarction." In such a tower or Babel, a clinician using "MI" as a search term, whether on the World Wide Web or hospital information system, could miss many medical records containing the concept myocardial infarction. Even worse, the clinician might actually be using "MI" to search for mitral insufficiency. Without institutional, specialtyspecific , regional, or national standards for vocabulary use, seemingly minor variations in how clinical...