Studies show that artificial intelligence tools have the potential to improve patient health literacy

An example ComprehE Notes test question with embedded NoteAid definitions implemented in a web application. In this example, the definition of “ferritin” (gray box) helps to understand that the bold text describes the blood iron level test. Credit: DOI: 10.2196 / 26354

A federal regulation requiring healthcare providers to provide patients with free, convenient and secure electronic access to their personal medical records came into effect earlier this year. However, even if you provide patients with access to clinician notes, test results, progress documents, and other records, they will understand those records and make appropriate health decisions based on what they read. Preparations for lowering are not done automatically. “Medicalese” can stumble even the most educated laymen, and studies have shown that low health literacy is associated with poor health.

John Luller, a researcher at the University of Notre Dame and an assistant professor of information technology, analysis and operations at Mendoza College of Business, is part of a team working on web-based natural language processing systems. health Patient literacy to access their records through the patient portal. A project based at the University of Massachusetts Amherst, NoteAid conveniently translates medical terms for healthcare consumers.

Lalor has worked with the team to develop Compreh ENotes, a tool that specifically assesses the understanding of electronic health record (EHR) notes. They also use crowdsource workers to make active interventions like NoteAid, which automatically define medical terms, make passive systems like MedlinePlus available on the Web. We compared how they improved their patients’ EHR literacy. The study found that NoteAid significantly improved health literacy scores compared to those without resources and those with access to MedlinePlus.

“Both of these studies used Amazon Mechanical Turk crowdsourced workers, and demographic groups of participants were generally associated with reduced health literacy demographic groups (eg, older and educated). It turns out that it doesn’t overlap well with the low people), “says Lalor. .. “In this study, we wanted to see if the definition tool, NoteAid, would work for real patients in the hospital.”

Their latest research was published in the May issue of Journal of Medical Internet Research, The team recruited 174 people waiting for an appointment at a regional hospital in Massachusetts.

Participants were presented with either the NoteAid version of the ComprehE Notes test and either a definition of a medical term that can be displayed by hovering the mouse over the text, or an undefined version.

“We assumed that the NoteAid tool would actually improve the performance of our understanding device, and that’s what it did,” Lalor said. We also found that the average score of hospital participants was significantly lower than the average score of crowdsourced participants. This was consistent with the low level of education of the regional hospital samples and the overall impact of the level of education on test results.

Lalor explained that these findings are important for several reasons. “First, by demonstrating that NoteAid is effective for local patients, we can generalize its usefulness to real patients beyond crowdsource workers,” he said. “The same is true for testing EHR memo comprehension. Both are related to recent legislation requiring patients access to EHR, including memos.”

With evidence that natural language processing tools can significantly improve patient health literacy, Lalor said the tools from both the physician’s perspective on accuracy and the patient’s perspective on reading level. He states that he is working on evaluating and improving the dictionaries used by. .. He also said: “The last part is a higher level question about what should be included in the dictionary as jargon and what is just a rare term, or something that may not be understandable but important. Note.”

Defining all the words in a medical record can overwhelm the patient. “If they know they need a particular term, they may be more likely to read them, internalize them, and better understand the notes,” he said.

At the undergraduate level, Lalor teaches unstructured data analysis courses. He also teaches in the Master of Science in Mendoza’s Business Analysis Program.His research interests are machine learning Natural language Processing related to model evaluation, quantification of uncertainty, interpretability of models, and application to biomedical informatics in particular.

“Assessment of NoteAid Efficacy in Community Hospitals: Randomized Trials of Electronic Carte Memo Understanding Interventions in Patients” by Wen Hu, Matthew Tran, Kathleen Mazor, Hong Yu, and Vanderbilt University at the University of Massachusetts.

Universal Health Literacy Precautionary Measures Recommended

For more information:
John P Lalor et al, Assessment of NoteAid Efficacy in Community Hospital Settings: Randomized Controlled Intervention for Understanding Electronic Health Record Memo in Patients, Journal of Medical Internet Research (2021). DOI: 10.2196 / 26354

Quote: Artificial intelligence tools may improve patient health literacy, according to studies (26 July 2021) 26 July 2021 Obtained from intelligence-tool-patient-health.html

This document is subject to copyright. No part may be reproduced without written permission, except for fair transactions for personal investigation or research purposes. The content is provided for informational purposes only.

Studies show that artificial intelligence tools have the potential to improve patient health literacy

Source link Studies show that artificial intelligence tools have the potential to improve patient health literacy

Related Articles

Back to top button