Speech tools to spot hospital or ER risk during home nursing visits
Official title: Using automated speech processing to improve identification of risk for hospitalizations and emergency department visits in home healthcare
Funded at Columbia University Health Sciences · NIH-11370222 · Data updated 2026-11-12
The project uses computer analysis of nurse-patient conversations during home healthcare to find people at higher risk of hospitalization or emergency department visits.
Quick facts
| Grant type | R01 grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | Columbia University Health Sciences NIH-funded |
| Lab location | 1 site (New York, United States) |
| Project start | June 15, 2023 |
| Project end | February 28, 2027 |
| Funding year | 2026 |
| Award (this year) | $580,605 |
| NIH project number | 5R01AG081928-04 |
| Project ID | NIH-11370222 on NIH RePORTER |
Who could benefit from this research
✓Could help
People receiving in-home nursing care from participating home healthcare agencies, especially older adults, would be the ideal candidates.
✕Less likely to help
People who do not receive home nursing visits or who cannot or will not allow audio recording of their visits are unlikely to benefit directly from this work.
What this research studies
Researchers will capture and process audio from nurse-patient conversations during in-home nursing visits and use automated speech and language analysis to pull out risk-related cues. They will refine a speech-processing system to identify words, tone, and communication patterns linked to future hospital or ER use. The team will combine these speech-derived signals with electronic health record data to see if prediction of hospitalizations and ED visits improves. Findings will inform whether adding conversation data could help care teams spot problems earlier.
Why it matters
Potential benefit: If successful, this could help nurses and care teams detect worsening problems sooner and reduce preventable hospital and emergency visits.
How similar studies have gone: Automated speech and language analysis has shown promise in other health areas, but applying it to predict hospital and ER risk in home healthcare is novel and not yet widely proven.
Where this research is happening
-
Columbia University Health Sciences
New York, United States Active
Researchers
- Principal investigator: Topaz, Maxim — Columbia University Health Sciences
- Study coordinator: Topaz, Maxim
Related research
Other NIH-funded projects on the same topics.
-
Silent speech interface using a wearable tongue tracker to help people after laryngectomy communicate
A wearable system with tongue sensors and intraoral components aims to let people without a vocal cord use natural-sounding speech in daily life.
-
Overlap between how we speak and how we hold words in memory
A Duke-led project studies how the brain that talks relates to keeping verbal information in working memory, using advanced brain recordings from patients already undergoing medical care.
-
AI-powered system to improve teamwork in cardiac surgery
A research project is testing an AI-driven system that listens to team conversations during heart surgeries to spot teamwork patterns and give real-time feedback to improve patient safety.
-
Caregiver Speech and Infant Brain Development in Siblings of Children with Autism
This project follows babies who have an older sibling with autism to explore how caregiver talk and everyday interactions relate to early brain and behavior development.
-
How speech and swallowing work in people with Down syndrome
Researchers will look at how differences in the mouth, throat, and brain affect speech clarity and safe swallowing in children and adults with Down syndrome.
-
Information and social media use for autism info in young children
A research project looks at how speech-language pathologists and parents of young autistic children use social media for information and what makes that information trustworthy.
About this research
- This is an active NIH-funded research project — typically early-stage science, not a clinical trial accepting patient enrollment.
- Some NIH-funded labs run parallel clinical studies or seek volunteers for related work. To check, contact the principal investigator or institution listed above.
- For full project details, budget, and progress reports, visit the official NIH RePORTER page.