Inside Angle

Can the use of language on social media reveal information about a patient's mental health? Can human language, with all its ambiguities and complexities, be analyzed to identify behavioral issues? And what's the boundary, ethically, of tapping into language sources? In this episode, Dr. Gordon Moore speaks with Philip Resnik, Professor of Linguistics at the University of Maryland, about the intersection of machine learning, natural language processing and mental health. Resnik details the work he and other researchers are doing to research human language connections by building a secure cloud enclave of language data where scientists are granted access, allowing them to collaborate and work at a scale that, until now, has been unimaginable.

How much time does it actually take for primary care physicians and specialists to deliver care, and should time be considered when determining payment? When it comes to physician performance, are we measuring what's truly important or just analyzing quality data that's readily available? Will new CMS rules governing E/M coding help or hinder the practice of medicine? In this episode, Dr. Gordon Moore talks with self-proclaimed healthcare contrarian Dr. Robert Berenson, a Fellow with the Urban Institute, about how to reform our current inefficient healthcare delivery system.

Atrium Health's Dr. Geoffrey Rose and Heather Joyner describe their work with clinicians to find a common language for documentation in the medical record. The result? Better quality scores and fewer denials.

Imagine having to pay for a car piece by piece, shopping for a steering wheel, brake system and transmission all as separate purchases. Not a great idea, right? Now imagine shopping for health care that way, down to every doctor visit, lab test or procedure. Yet, that describes the many confusing and complex choices facing U.S. consumers as they try to manage their health care. In this episode, Mike Fay, Vice President for Health Networks and Innovation at Wellmark, describes what happens when health plans collaborate with providers to offer consumers more transparency about healthcare quality and pricing.

What if clinicians could use technology to find indicators of cognitive impairment hidden in the medical record? Dr. Andrea Gilmore-Bykovskyi of UW-Madison explains how EMR free text can be used to identify patients with early signs of dementia.

In the mid-1980s, Dr. David Buck developed clinics for the indigent population of Houston, Texas where he learned first-hand that his patients needed social and behavioral supports outside of the traditional medical domain. His work led him to join forces with data scientist Kallol Mahata and together they founded Houston's Patient Care Intervention Center (PCIC) to improve care for the city's vulnerable populations. Buck and Mahata developed a technology platform that integrates social data – from jails, shelters, food banks and other community service organizations – with medical data to provide a holistic view of the patient. Always based on the patient's values and health goals, their work to integrate social and behavioral data with care coordination has resulted in a savings of over $5 million and a reduction in ER visits by 54 percent.

If quality outcomes for length of stay, readmissions and other measures are higher than industry norms, is it a question of poor clinical care? When Dr. Gene Christian, CMO for St. Mary's Hospital, Bon Secours discovered that physicians weren't documenting completely and accurately in the medical record, he became a disciple of clinical documentation improvement.

At age 31, when Dr. Alan Glaseroff learned he had Type 1 diabetes, he began collaborating with his diabetes patients on new ways of self-managing the condition. His pioneering work in patient engagement led him to Stanford University, where today he helps individuals with multiple chronic conditions believe it's not where you start, but where you end up that matters.

Are we at a tipping point in making the move from volume to value? Does your organization have the right mindset to be successful? Mark Sonneborn of the Minnesota Hospital Association discusses new data initiatives that help providers better understand risk and reduce avoidable care.

By 2020, medical knowledge will double every 73 days. Can machine learning and natural language processing (NLP) help clinicians drowning in information? Dr. John Cromwell of the University of Iowa discusses his efforts to bring these technologies to the frontlines of medicine.