AI improves electronic health record (EHR) systems

Source: geralt / pixabay

Researchers from the Massachusetts Institute of Technology (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) and Beth Israel Deaconess Medical Center have teamed up to improve electronic health records (EHRs) with machine learning from artificial intelligence ( IA) and published their results in a recent to study.

Automation of patient health records gives hope to clinicians, patients and stakeholders, such as increasing data transfer speed, reducing paper record maintenance costs, increasing efficiency, improving results by avoiding or reducing clinical errors. However, electronic health records have yet to achieve many of these positive benefits and are a major cause of burnout and stress among physicians according to researchers. Clinicians spend time using electronic health records instead of talking with patients.

The global electronic health records market was US $ 26.8 billion in 2020, with North America holding the highest revenue share of 45% according to Grand View Research. The global EHR market is expected to grow at a compound annual growth rate (CAGR) of 3.7% between 2021 and 2028, according to the same report. Leading companies in the EHR market include Epic Systems Corporation, NextGen McKesson Corporation, MEDITECH, Allscripts and Cerner.

Patient health records were largely in paper form until electronic health records (EHRs) first appeared in the 1960s. Electronic health records, also known as electronic medical records (EMRs) , are used in medical settings to electronically store patient information such as medical history, prescriptions, lab test results, x-ray images, demographics, immunization status, billing history and more of data.

Fast forward to modern times and electronic health records have almost completely replaced paper records. Among U.S. office-based physicians, 85.9% use an electronic medical record system, according to figures from the Nation Center for Health Statistics of the Centers for Disease Control and Prevention (CDC).

According to the researchers, the process of clinical documentation remains a “tedious, lengthy and error-prone process.” Scientists cite how this is especially the case in emergency rooms, where clinicians can see up to thirty-five patients during a shift, forcing them to quickly absorb the contents of the medical history. patients from “multifaceted requirements and fragmented interfaces for exploring and documenting information” that are often new to them before creating an informed diagnosis and focused care plan.

Although EHRs offer vast improvements in speeding up access and retrieval of patient records, documentation systems can be time consuming and cumbersome for clinicians to use.

“To better support this synthesis of information, clinical documentation tools must allow rapid contextual access to the patient’s medical record,” write the researchers.

To solve these problems with electronic health records, researchers at MIT CSAIL created a machine learning system called MedKnowts and implemented it at Beth Israel Deaconess Medical Center in Boston, Massachusetts. The AI-backed EHR system integrates the information retrieval system with a note-taking editor so that the search is efficient. The system enables clinicians to use natural language and automates the entry of structured data. Documentation is streamlined with features like autofill text, proactive information retrieval, and easy parsing of long notes.

According to the study, the average score on the scale for using the system by scribes was 83.75 out of a possible 100. Researchers report that scribes found their AI-based system easy to learn and use, and would use it frequently.

With this new proof of concept, the researchers plan to further improve AI machine learning to identify the part of a patient’s health record that is most relevant for the clinician to focus on reading. The researchers plan to incorporate contributions from clinicians, such as medical terminology, in the future so that the system adapts over time. In the next steps, the team is exploring the possibility of commercializing AI machine learning technology in the future.

Copyright © 2021 Cami Rosso All rights reserved.


Source link

Comments are closed.