Credit score: Pixabay/CC0 Public Area
Synthetic intelligence (AI) voice brokers helped older adults with hypertension to precisely report their blood stress readings and improved blood stress administration, in response to preliminary analysis offered on the American Coronary heart Affiliation’s Hypertension Scientific Classes 2025 held in Baltimore, September 4–7, 2025.
“Controlling blood pressure remains a cornerstone for improving cardiovascular outcomes for patients, however, capturing timely, compliant blood pressure readings remains a challenge, particularly for patients with limited access to care,” stated lead examine creator Tina-Ann Kerr Thompson, M.D., senior vice chairman of the first care service line and government director of the inhabitants well being collaborative at Emory Healthcare in Atlanta.
“In our study, we were able to improve accuracy of blood pressure measures and patient outcomes.”
AI voice brokers are conversational techniques powered by giant language fashions that may perceive and produce pure speech in actual time when interacting with people.
This examine included 2,000 adults, the bulk aged 65 and older, and was designed to judge the effectiveness and scalability of a voice-enabled AI agent in participating sufferers to self-report correct blood stress readings, rather than a telephone name with a well being care skilled about their blood stress measures.
The AI agent additionally recognized sufferers in want of follow-up medical care primarily based on their blood stress readings.
The AI voice-agent calls to sufferers have been made utilizing commercially accessible AI in a number of languages, together with English and Spanish. A blood stress studying outdoors the edge vary for readings that change primarily based on the presence of different circumstances, akin to diabetes, resulted within the name being escalated to a licensed nurse or medical assistant.
The presence of signs akin to dizziness, blurred imaginative and prescient or chest ache additionally prompted escalation of the decision. Escalation to extra care was rapid in pressing conditions or inside 24 hours for non-urgent points.
The sufferers have been contacted by the voice agent to offer latest blood stress readings or to conduct stay measurements in the course of the name. After the decision, the readings have been entered into the affected person’s digital well being file and reviewed by a clinician. Name routing and referrals for care administration have been prompted for sufferers with difficult-to-control hypertension.
This course of decreased the guide workload by clinicians and resulted in an 88.7% decrease cost-per-reading. This quantity was calculated by evaluating the price of commercially accessible AI voice brokers with using human nurses to carry out related duties that lead to efficiently acquiring affected person self-reported blood stress readings.
The examine discovered that integrating AI into scientific workflows lowered prices and improved care administration for sufferers. Through the examine interval:
85% of sufferers have been efficiently reached by the voice-based AI agent.
Of these sufferers, 67% accomplished the decision, and 60% took a compliant blood stress studying in the course of the name. Amongst these sufferers, 68% met CBP (controlling blood stress) Stars compliance thresholds.
General, 1,939 CBP gaps have been closed, elevating the measure from 1-Star to 4-Star efficiency—a 17% enchancment. The Medicare Benefit (MA) and Healthcare Effectiveness Knowledge and Info Set (HEDIS) CBP measure elevated from a beforehand reported 1-star score to 4-star score.
On the finish of every accomplished name, sufferers obtained a two-question survey to fee their satisfaction on a scale of 1 to 10, with 10 being 100% glad. Among the many accomplished calls, the common patient-reported satisfaction fee exceeded 9 out of 10, reflecting a wonderful total expertise with the voice-based AI agent.
“We were surprised by the high patient satisfaction scores after interacting with artificial intelligence-based voice agents,” stated Thompson. “We are excited for what that means for the future, since patient engagement and satisfaction are so critical to health care outcomes.”
“This could be a game-changing study,” stated Eugene Yang, M.D., M.S., FACC, an American Coronary heart Affiliation volunteer knowledgeable.
“Accurate blood pressure readings are essential to improving control, and new approaches can help make that possible. Breakthrough AI technologies like this could transform how we manage blood pressure by reaching patients wherever they are and addressing critical barriers, such as limited access to care and gaps in patient support.”
Yang, who was not concerned on this examine, is a professor within the division of cardiology and the Carl and Renée Behnke Endowed Chair for Asian Well being at College of Washington College of Drugs.
The examine has a number of limitations. This examine was observational and didn’t have a management group. The consecutive AI calls weren’t in comparison with human calls; as an alternative, AI voice calls have been deployed as a result of it was not attainable to make an satisfactory variety of human-only calls. As well as, the examine was retrospective, that means it reviewed present knowledge, and analysis was accomplished after the clinically recognized calls have been already made.
Examine particulars, background and design:
Individuals included 2,000 adults; a majority have been ages 65 or older (common age of 72 years; 61% ladies) receiving take care of hypertension.
Evaluation of digital well being data recognized sufferers who have been lacking blood stress knowledge or whose most up-to-date BP studying was not throughout the regular vary of
The examine was carried out with sufferers at Emory Healthcare in Atlanta throughout a 10-week interval. Sufferers obtained a minimum of one telephone name in the course of the examine. Sufferers obtained multiple name if they didn’t reply the telephone.
Sufferers with open gaps in managing blood stress have been recognized by means of digital medical data (EMR) and payer analytics. Affected person lists have been reviewed to make sure the knowledge of their data was appropriate, they usually have been verified for outreach by a scientific operations crew to make sure real-time accuracy of gaps earlier than outreach to the sufferers.
AI texts, telephone calls from the traditional care crew, latest scientific visits the place documentation may very well be discovered for a blood stress studying and generative AI voice brokers have been used to contact sufferers to offer latest blood stress readings or take their blood stress studying in the course of the name. These included any latest scientific visits the place documentation may very well be discovered for a BP recorded.
A post-call validation step was built-in into the workflow, through which readings have been entered into the EHR, reviewed by a clinician and submitted as supplemental knowledge to shut the Stars high quality hole. For sufferers with uncontrolled hypertension, scientific escalation referrals have been made to care administration groups.
The Facilities for Medicare and Medicaid Providers (CMS) developed the Star Rankings system, referred to as MA Stars, to fee Medicare Benefit (MA) (Half C) and prescription drug (Half D) plans on a 5-star scale with 1 being the bottom rating and 5 being the best rating. MA plans are plans from personal insurance coverage corporations authorized by Medicare and never issued by Medicare itself. Hospitals, care facilities and clinicians are eligible to obtain a bonus fee enhance in the event that they obtain a minimum of a 4-star score.
Self-measured blood stress is a spotlight space of Goal:BP, an American Coronary heart Affiliation initiative that helps well being care organizations enhance blood stress management charges by means of an evidence-based program. Dwelling blood stress monitoring is beneficial for all adults with any stage of hypertension, as famous within the Affiliation’s new 2025 guideline on hypertension, launched in August.
Supplied by
American Coronary heart Affiliation
Quotation:
AI helps older adults report correct blood stress readings at dwelling (2025, September 7)
retrieved 7 September 2025
from https://medicalxpress.com/information/2025-09-ai-older-adults-accurate-blood.html
This doc is topic to copyright. Other than any truthful dealing for the aim of personal examine or analysis, no
half could also be reproduced with out the written permission. The content material is supplied for info functions solely.

