Future Bet: From Predictive to Proactive Care with AI
VISION & GOALS
AI, artificial intelligence, is a term that has become ubiquitous with advancements in technology and synonymous with innovation.
At Houston Methodist, our vision with employing AI technology is to be proactive rather than reactive and enhance human touch by leverage. We want to use AI data not to share what happened in the past or what is happening at present, but instead to make future predictions for better patient care.
By employing AI technology, our goals are to:
-
Increase productivity – unburden clinicians from documentation
-
Promote hands-on patient care – free up doctors and nurses to focus on the patient
-
Decrease mortality – patients receive more targeted care to their specific condition
-
Improve patient quality of life post-discharge – clinicians can provide a better plan of care for their patients
OUR SOLUTION
We understand the need for partnerships (large companies with lots of resources), expertise (humans to train the AI and oversee the small percentage of errors), and cloud computing and storage (moving everything into the cloud). We also place the highest priority on AI implementation following the most comprehensive degree of safety and security measures.
CURRENT RESULTS & PARTNERSHIPS
AI tools are helping us tackle the continuum of care. We are able to connect with and follow a patient’s health care journey in ways we have never been able to before by categorizing unstructured data and presenting it in a useful way that provides meaningful outcomes. Additionally, a completely seamless integration for both patients and providers is one of the best aspects of this new technology implementation.
One of the most impactful AI implementations is our Virtual ICU (vICU). In partnership with MIC Sickbay, Houston Methodist clinicians developed over 20 algorithms to provide fast-paced, immediate ICU level care to patients via virtual nurses. Initial outcomes showed codes decreased 37% by monitoring virtually and catching potential issues earlier.
Another pilot initiative with cardiovascular and orthopedic surgery patients. Health Data Analytics Institute (HDAI) leverages a large database of Medicare data to generate personalized information for every patient, including likelihood to develop a chronic condition or risk of an adverse event.
Similarly, we are engaging in a pilot program is with Pieces, which uses natural language processing to pull important insights from clinical notes in the Electronic Health Record (EHR). We are setting up a process to integrate these insights seamlessly into our Epic interface and reduce length of stay and readmission rates.
Through these new models, we are also able to predict better population health data and link to stronger community benefits programs. We appreciate working with companies that can use our partnership as a means to improve their products.
NEXT STEPS
Each initiative is rolled out initially only to select service lines in order to prove the impact. This model allows us to be agile and “succeed fast, fail fast” with innovation projects. With the proven success of several AI initiatives, the next steps are to implement comprehensively at one hospital and then expand across the entire hospital system to reach all clinicians, staff and patients.