Quire Granted Patent on Text-Powered, Predictive Modeling for Healthcare


The United States Patent and Trademark Office has granted Quire, Inc, a patent on the companys method for using text in medical records to predict health outcomes, risks, and behaviors. The unique method uses practitioner observations, captured in clinical notes, to build a patients story which can then be used to identify and prioritize patients with similar stories for outreach and interventions.

Text is the easiest data to manage, contains the richest information on patients, but can be a challenging data-type from which to extract actionable information. Quires capability enables practitioners to quickly build, test, and implement predictive models using only text, said Quire CEO Brad Silver.

For Brigadier General (Ret) Stephen N. Xenakis, M.D., Quire allows him to scale his knowledge to assess hundreds of thousands of people, fast. In my areas of interest “ risk prediction for suicide and dangerous behavior “ only 53% of the people at risk for harm have self-selected for treatment or support. However, 100% of the people with serious problems have clinical signals, recorded in notes, like insomnia, headaches, alcohol use, relationship problems etc. recognized as patterns alerting practitioners to be concerned.

With value-based care, the largest portion of immediate savings opportunities are related to proactive management of patients with chronic conditions. Provider notes offer the most detailed and nuanced understanding of patients and their propensity to engage, or not, in behaviors which increase the risk of emergency room visits or inpatient admissions.

Social and behavioral determinants of health are a hot topic with providers. Many health systems are spending tons of time and money to implement assessment templates, which will be inconsistently used, to collect this information. However, this data already exists, because the people treating patients record things like, ˜transportation is a problem, no family support, cant afford medications, etc. Quire uses this information in notes to predict who should be prioritized for outreach, said Silver.

Text also enables practitioners and operators to unpack drivers of poor outcomes and make workflow adjustments to address those drivers in clinical process. A big challenge with AI and deep learning approaches is that they are still a black box. These methods can produce good predictions, but they dont reveal why. With Quires unique approach to predicting outcomes, the underlying basis for the predictions can be more readily teased out of the text, explained Ramin Homayouni, Ph.D., Quires chief scientist and current director of Population Health Informatics at the Oakland University William Beaumont School of Medicine.

Silver added, Quire gives providers the ability to leverage, at a population-level, the best means of recording interactions with patients, their notes.


Quire provides software and services needed to extract actionable information from text. With a focus on healthcare, the company also provides capabilities to law enforcement and other markets where text contains data needed for decision-making.

Brad Silver | [email protected]
| 901.866.1624