Official websites use .gov

A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS

A lock ( 🔒 ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Need Help

HEAR-HEARTFELT (Identifying the risk of Hospitalizations or Emergency depARtment visits for patients with HEART Failure in managed long-term care through vErbaL communicaTion)

Award Number: R00HL169940
ORGANIZATION: NATIONAL HEART, LUNG, & BLOOD INSTITUTE
OPDIV: NIH
AWARD CLASS: DISCRETIONARY
AWARD ACTIVITY TYPE: SCIENTIFIC/HEALTH RESEARCH (INCLUDES SURVEYS)
PERIOD OF PERFORMANCE START DATE: 08/20/2026
PERIOD OF PERFORMANCE END DATE: 06/30/2029

Group Awards By:

View Award Description

HEAR-HEARTFELT (Identifying the risk of Hospitalizations or Emergency depARtment visits for patients with HEART Failure in managed long-term care through vErbaL communicaTion) - As a result of efforts to reduce healthcare costs and improve quality, patients with heart failure are increasingly receiving treatment in community-based programs, such as managed long-term care, which supports older adults in remaining independent. These programs aim to reduce unplanned hospitalizations and emergency department (ED) visits, yet heart failure remains a leading cause of such events. In this setting, care coordinators (i.e., registered nurses or social workers) maintain regular telephone contact with patients to ensure care aligns with medical needs. From a linguistic perspective, verbal communications between patients and providers are information-seeking and problem-focused. From an acoustic perspective, heart failure can affect speech production due to fluid retention, fatigue, or laryngeal nerve compression. Although these conversations contain rich, temporally sensitive information about patient status, they are largely untapped in risk stratification models. To address this gap, we aim to evaluate whether audio-recorded telephone communication can be used to improve the prediction of hospitalization and ED visits. During the K99 phase, we successfully identified clinically relevant information embedded in patient–coordinator conversations and extracted risk factors using three complementary approaches: (1) analyzing conversational characteristics to capture interaction patterns; (2) developing linguistic indicators based on a curated lexicon of risk-related language, including heart failure symptoms, poor self-management, and psychosocial stressors; and (3) extracting acoustic features from voice signals. The proposed R00 phase builds directly on this foundation by advancing to predictive modeling with these multimodal features and expanding the data collection period. In R00 Aim 1, we will evaluate the effectiveness of large language models (LLMs) in identifying risk factors and communication dynamics from transcribed conversations, building on infrastructure and NLP models developed during the K99 phase. R00 Aim 2 focuses on developing time-aware machine learning algorithms for longitudinal risk prediction. We will construct and evaluate models using integrated data from (a) structured electronic health records, (b) care coordination notes, and (c) verbal communication features. We will assess whether combining these diverse data sources improves the predictive performance of risk models and enhances early detection of clinical deterioration in this high-risk population, ultimately informing real-time


Issue Date FY	Funding FY	Legal Entity Name	Legal Entity Address	Legal Entity City	Legal Entity State	Legal Entity Zip Code	Legal Entity COUNTY	Legal Entity COUNTRY	Assistance Listing	Award Code	Budget Year	Action Date	Action Type	Action Amount

Issue Date FY: 2026 ( Subtotal = $237,153 )
2026	2026	TRUSTEES OF THE UNIVERSITY OF PENNSYLVANIA, THE	3451 WALNUT ST	PHILADELPHIA	PA	19104	PHILADELPHIA	USA	Cardiovascular Diseases Research	000	3	8/21/2026	EXTENSION WITH OR WITHOUT FUNDS	$237,153
														Subtotal = $237,153

Grand Total All Awards = $237,153

Top

All Categories

About

Search

Reports

Data Submission

Award Information

HEAR-HEARTFELT (Identifying the risk of Hospitalizations or Emergency depARtment visits for patients with HEART Failure in managed long-term care through vErbaL communicaTion)

Award Number: R00HL169940

ORGANIZATION: NATIONAL HEART, LUNG, & BLOOD INSTITUTE

OPDIV: NIH

AWARD CLASS: DISCRETIONARY

AWARD ACTIVITY TYPE: SCIENTIFIC/HEALTH RESEARCH (INCLUDES SURVEYS)

PERIOD OF PERFORMANCE START DATE: 08/20/2026

PERIOD OF PERFORMANCE END DATE: 06/30/2029

Federal Websites

Department of Health & Human Services

HHS Operating Divisions

HHS Staff Divisions

Download A Document Viewer