Here’s a telling statistic: Hospital readmissions cost Medicare, Medicaid and self-insured employers $180 billion each year. Why are we spending so much on readmissions? More importantly, why are so many people being readmitted? Unfortunately, many of the processes in place to help patients confidently and safely continue recovering at home are fragmented and broken.
When a person is admitted to the hospital, their care team is focused on diagnosing and treating the problem that led to the admission. But in-hospital treatment isn’t the whole story. Often people leave the hospital without having a good sense of where exactly they are in their own recovery or how to continue treatment outside the hospital. Many are ill-equipped and ill-prepared for their continued recovery at home.
Hospitalizations disrupt a person’s physical and emotional stability and leaving the hospital can be a time of acute anxiety and vulnerability. Entering this vulnerable period alone with little to no preparation can lead to a slew of challenges, resulting in what some doctors call “post-hospital syndrome.” During this period, people are at increased risk for setbacks that can lead to readmissions, worsening health and death.
Recovery coaches and augmented intelligence
One way to address this problem is to employ recovery coaches, behavioral health coaches who specialize in transition of care. They’re experts in supporting an individual’s unique behavioral, social, environmental and medical needs, an approach known as “whole person care.”
A recovery coach works one-on-one with a person who has experienced a big health change such as a hospitalization or a new diagnosis. Recovery coaches can support a person’s emotional journey, help them with day-to-day activities related to their recovery and motivate them to modify their behavior for better recovery.
Democratizing recovery coaching programs can be a gamechanger for improving health outcomes. To do this, we need programs that are affordable, high-quality and personalized. This is where technology can help.
Medical technology innovations are moving at an astounding rate. Analytics allow providers to spot trends that might otherwise go unnoticed; data enables efficient and effective support through a personalized approach; and artificial intelligence such as natural language processing, voice biomarkers and predictive algorithms can save time and improve care.
Many healthcare organizations are already using AI-based solutions for problems that range from medication non-adherence to cost management issues. New concepts and uses for AI are emerging every day in the field.
AI algorithms built to provide what’s called “augmented intelligence” can be particularly beneficial. The AMA uses the term augmented intelligence in referring to “artificial intelligence that focuses on AI’s assistive role, emphasizing that its design enhances human intelligence rather than replaces it.” Already, this concept has been applied to create AI algorithms that look for patterns in how patients speak and to create meaningful notes for clinicians, freeing them to focus more on establishing real human connection.
On the horizon are AI technologies that will help diagnose a variety of conditions like depression or heart failure. Companies like Headspace have begun to build AI algorithms to augment their behavioral health coaches and personalize the experience they provide.
For recovery coaching programs, augmentation offers possibilities like creating automated mental health screening tools, identifying gaps and risks specific to individual members and personalizing education to the member’s learning style. All of these tools will help increase the frequency of early detection and effective intervention.
Today’s recovery journey is broken, but recovery coaching programs bolstered by augmented intelligence can drastically reduce readmission rates, as well as improve overall health and patient experience. As coaching programs grow in popularity, the question remains about how to best scale them. Augmented intelligence is making it possible.
