Clinical Informatics for Emergency Physicians: Part 1 – A Brief History

Gregory Han, MD
Ali Abdelati, MD
Katherine Godfrey, MD
Christopher Reisig, MD
Michael Alfonzo, MD

Editor’s Note
Clinical informatics increasingly shapes how emergency physicians deliver care, from electronic health records and clinical decision support to emerging applications of artificial intelligence. This article is the first in a short series aimed at emergency physicians who are curious about informatics and its role in modern emergency care. We begin with a brief history of how clinical informatics evolved alongside healthcare technology, setting the foundation for future pieces that will explore operational informatics, clinical decision support optimization, alert fatigue, and the growing role of generative AI in the emergency department.

Introduction

Clinical Informatics sits at the intersection between frontline clinical care, technology, and operations. For emergency physicians encountering informatics every shift, understanding what informaticians do begins with understanding how the field evolved alongside modern healthcare technologies.

A Brief History of Electronic Health Records

Computers first entered use in healthcare in the 1960s in administrative and billing roles. While clinical workflows have historically been slower to change, some hospitals began to experiment with computerized clinical tools shortly thereafter. One of the first examples of a true Electronic Health Record (EHR) was the HELP system, which was first developed at LDS Hospital in Salt Lake City circa 1967. It began as a clinical database that supported LDS Hospital’s cath lab and cardiac critical care unit with rule-based alerts based on lab, pharmacy, and vitals data. It quickly grew to support clinical operations hospital-wide in the 1970s, and across several hospital consortiums by 1999 . Around the same time, Massachusetts General Hospital (MGH) piloted its COSTAR system, which stored digital charts for Harvard Community Health Plan ambulatory clinics. COSTAR was designed modularly, allowing its architecture to be implemented elsewhere in parts as desired by other MGH affiliates or other hospital systems.

Most early EHRs grew bottom-up from certain units of a hospital before expanding outwards. The Veterans Affairs (VA) system did things differently. In 1977, it launched the Decentralized Hospital Computer Program (DHCP). This system enabled data sharing across the VA’s nationwide network of hospitals and ambulatory clinics, which was a remarkably early example of an interoperable EHR. Commercial EHR vendors soon emerged as well. Human Services Computing Inc. (later, Epic) was founded in 1979, while PGI & Associates developed what would become Cerner in the same year. Allscripts followed shortly thereafter in 1986.

Many of these EHRs successfully found market share throughout the US. Hospital systems started phasing out paper charting in favor of computers throughout the next few decades. However, EHRs and computers did not become widespread until the turn of the millennium, where high-profile patient safety events accelerated the push towards electronic clinical workflows. For example, the Libby Zion case (where two serotonergic drugs were verbally ordered for a patient already taking an MAOi), was held up as an example of where computerized drug interaction safeguards may have prevented patient harm.

Major reports like To Err is Human and Crossing the Quality Chasm called for the implementation of clinical decision support (CDS) tools that could help clinicians spot errors before they happen. The United States Government saw the potential of EHRs to provide said CDS tools, so it encouraged health systems to adopt EHRs through the HITECH Act. This pushed health systems to adopt EHRs over several Meaningful Use stages in the 2010s, with penalties for those who didn’t comply in time. Stage 1 included e-prescribing lists of allergies, problems, vitals, and labs, which reflected the early focus on preventing harm. Later stages focused on communication (e.g. patient portals and secure chat) and interoperability between systems, allowing for easier patient data sharing to promote clean transitions of care. Today, one might recognize these features as everyday parts of their practice. However, even 20-30 years ago, many physicians were still charting with pen and paper.

The Evolution of Clinical Informatics

Early computerized clinical systems were often developed by interdisciplinary teams of computer scientists and clinicians. COSTAR, for example, was built by Octo Barnett and Jerome Grossman from the Laboratory of Computer Science at MGH. One of Dr. Barnett’s students, Ted Shortliffe, wrote his dissertation on the MYCIN system, which was designed to help clinicians select the right antibiotic for a given disease. This was an example of a CDS tool decades before the nation would push for their widespread implementation, and Dr. Shortliffe himself would go on to become a founding father of the field of clinical informatics as both a physician and a computer scientist.

As work like this proliferated, departments of biomedical informatics emerged at several major academic medical centers. The University of Utah established one as early as 1964, while Columbia University, Harvard, and the Children’s Hospital of Philadelphia were just some of many who founded theirs throughout the rest of the 20^th century. Several professional societies of clinicians and computer scientist eventually consolidated into the American Medical Informatics Association (AMIA) in 1988, which has since been the national academic home for the specialty. It codified the core content for Clinical Informatics Fellowships in 2008. Clinical Informatics was recognized as a clinical subspecialty in 2011, and its board certification was finalized in 2013. Clinical informaticians have been formally trained and certified since, either by the fellowships that sprung up to meet the goals of the core content or through the practice pathway. The subspecialty is specialty-agnostic, leading to informaticians coming from all prior walks of life. Internists, pediatricians, pathologists, and surgeons have all become informaticians, as well as plenty of Emergency Physicians.

The Many Roles of the Informatician

Clinical informatics fellowships now span the country, befitting the breadth of the field. Many informaticians research applying computer technologies to the clinical space. Examples include using generative AI to enable ambient dictation and machine learning, which can extract additional meaning from existing information like signs of structural heart disease on an electrocardiogram. Some work in academic laboratories, while others work in industry.

Many other informaticians are deeply embedded in day hospital operations now that EHRs are ubiquitous. Most hospitals have dedicated informatics leadership with roles like the Chief Medical Information Officer (CMIO) and/or Chief Information Officer (CIO), who oversee domains ranging from quality improvement to cybersecurity. Department-level informaticians serve as the link between clinicians and their EHR vendor, optimizing clinical workflows by speaking the languages of the clinician and the computer analyst.

Emergency physicians encounter the work of informaticians every day. The rapid expansion of CDS tools through EHRs improved patient safety, especially surrounding patient allergies and drug-drug interactions. However, they also introduced an overabundance of interruptive alerts. This has led to the phenomenon of “alert fatigue,” where clinicians’ screens and cognitive load are oversaturated with alerts. A common focus of ED informatics has since been streamlining these alerts and reducing the number of interruptions.

Emergency Medicine and Clinical Informatics

There are many career paths within clinical informatics. Just like Emergency Medicine, it is a young field with a broad reach and many unique applications. Emergency physicians are also uniquely suited for informatics work. Informaticians need to serve the whole hospital system, and emergency physicians interact with nearly every division within the hospital while seeing patients with a staggering range of pathology and acuity. Shift work is also very compatible with scheduling an informatician’s day-to-day office work, yielding improved work-life balance.

Finally, informaticians play a key role in improving quality of care at scale, which resonates with the volumes of patients that emergency physicians see on a regular basis. For emergency physicians that are interested in shaping the future of care delivery across health systems, clinical informatics offers an increasingly powerful way to get involved.